8,557 results on '"LONG short-term memory"'
Search Results
2. State of the Art Recurrent Neural Network with Bidirectional Long Short-Term Memory for Cursive Handwriting Recognition
- Author
-
Jose, Manju, Udupi, Prakash Kumar, Li, Gang, Series Editor, Filipe, Joaquim, Series Editor, Ghosh, Ashish, Series Editor, Xu, Zhiwei, Series Editor, T., Shreekumar, editor, L., Dinesha, editor, and Rajesh, Sreeja, editor
- Published
- 2025
- Full Text
- View/download PDF
3. Ensemble Approach to Adaptable Behavior Cloning for a Fighting Game AI
- Author
-
García, José, Castro, Carlos, Valle, Carlos, Goos, Gerhard, Series Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Hernández-García, Ruber, editor, Barrientos, Ricardo J., editor, and Velastin, Sergio A., editor
- Published
- 2025
- Full Text
- View/download PDF
4. Deep Reinforcement Active Learning for Stress Recognition
- Author
-
Ngoc, Phan Anh, Nguyen, Ky Trung, Tran, Thanh-Tung, Jayatilake, Senerath, Nguyen, Thi Thanh Quynh, Li, Gang, Series Editor, Filipe, Joaquim, Series Editor, Ghosh, Ashish, Series Editor, Xu, Zhiwei, Series Editor, Thai-Nghe, Nguyen, editor, Do, Thanh-Nghi, editor, and Benferhat, Salem, editor
- Published
- 2025
- Full Text
- View/download PDF
5. Predicting Water Levels Using Gradient Boosting Regressor and LSTM Models: A Case Study of Lago de Chapala Dam
- Author
-
López-Barrios, Jesus Dassaef, de Anda-García, Ilse Karena, Jimenez-Cruz, Raul, Trejo, Luis A., Ochoa-Ruiz, Gilberto, Gonzalez-Mendoza, Miguel, Goos, Gerhard, Series Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Martínez-Villaseñor, Lourdes, editor, and Ochoa-Ruiz, Gilberto, editor
- Published
- 2025
- Full Text
- View/download PDF
6. A Fast and Accurate Reconstruction Method for Boiler Temperature Field Based on Inverse Distance Weight and Long Short-Term Memory
- Author
-
Huang, Rizhong, Zhang, Menghua, Li, Yichen, Huang, Ke, Huang, Weijie, Ghosh, Ashish, Editorial Board Member, Zhou, Lizhu, Editorial Board Member, Zhang, Haijun, editor, Li, Xianxian, editor, Hao, Tianyong, editor, Meng, Weizhi, editor, Wu, Zhou, editor, and He, Qian, editor
- Published
- 2025
- Full Text
- View/download PDF
7. A novel neural network architecture and cross-model transfer learning for multi-task autonomous driving
- Author
-
Li, Youwei and Qu, Jian
- Published
- 2024
- Full Text
- View/download PDF
8. Optimization of node deployment in underwater internet of things using novel adaptive long short‐term memory‐based egret swarm optimization algorithm.
- Author
-
Simon, Judy, Kapileswar, Nellore, Padmavathi, Baskaran, Devi, Krishnamoorthy Durga, and Kumar, Polasi Phani
- Subjects
- *
OPTIMIZATION algorithms , *MARINE resources , *INTERNET of things , *LONG short-term memory , *COGNITIVE learning , *NETWORK performance - Abstract
Summary: Optimizing node deployment in the underwater Internet of Things (UIoT) poses significant challenges due to the complex and dynamic nature of underwater environments. This research introduces the adaptive long short‐term memory‐based egret swarm optimization algorithm (ALSTM‐ESOA), a novel approach designed to enhance network coverage and performance efficiently. Unlike traditional methods, ALSTM‐ESOA incorporates cognitive learning capabilities from long short‐term memory (LSTM) and dynamic adaptation strategies inspired by the hunting behaviors of egrets. The algorithm's effectiveness was tested through extensive simulations in MATLAB, demonstrating notable improvements over existing models: network throughput increased by up to 55.56%, deployment time decreased by 88.89%, and energy efficiency improved significantly. These enhancements are critical for robust, real‐time data collection and monitoring in underwater settings, providing substantial benefits for marine research and resource management. The findings suggest that ALSTM‐ESOA significantly outperforms conventional algorithms, offering a promising new tool for the advancement of UIoT applications. After being implemented in MATLAB, the suggested ALSTM‐ESOA model for the node deployment optimization in UIoT is examined. The proposed ALSTM‐ESOA in terms of network throughput is 55.56%, 38.89%, 36.11%, and 11.11% better than CNN, LSTM, ARO‐RTP, and IGOR‐TSA, respectively. Similarly, the proposed ALSTM‐ESOA with respect to deployment time is 88.89%, 81.82%, 75%, and 50% better than CNN, LSTM, ARO‐RTP, and IGOR‐TSA, respectively. For the purpose of exploring marine resources, monitoring underwater environments, and conducting marine scientific investigation, the research's findings are extremely valuable. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
9. Application of long short-term memory network for wellbore trajectory prediction.
- Author
-
Huang, Meng, Zhou, Kai, Wang, Laizhi, and Zhou, Jianxin
- Subjects
- *
MATHEMATICAL formulas , *PREDICTION models , *GENERALIZED spaces , *CURVATURE , *AZIMUTH , *LONG short-term memory - Abstract
For improving the accuracy of wellbore trajectory prediction, especially in the build section, a new prediction model based on long short-term memory (LSTM) network was proposed. At the same time, the model was built by Python language and TensorFlow library. The well inclination and azimuth angle were predicted by the LSTM model. Moreover, the prediction accuracy of the LSTM was compared with that of the minimum curvature method. The results show that the average prediction error of the LSTM is much 50% lower than the minimum curvature method, and the proposed model in this article is more consistent with the actual drilling data. This method does not rely on any assumption of the path shapes and geometries. In addition, the model proposed in this article is easy to use and convenient for the engineering field application without the derivation of mathematical formulas. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
10. Pruning Long Short-Term Memory: A Model for Predicting the Stress–Strain Relationship of Normal and Lightweight Aggregate Concrete at Finite Temperature.
- Author
-
Dabbaghi, Farshad, Tanhadoust, Amin, and Ogunsanya, Ibrahim G.
- Abstract
While normal weight aggregate concrete (NWAC) can experience significant strength loss and spalling at high temperatures, lightweight aggregate concrete (LWAC) can maintain its structural integrity. Stress–strain relationship of concrete is an important test to perform during designing phase of concrete infrastructures. Therefore, this study focuses on exploring the stress–strain behavior of NWAC and LWAC under uniaxial compression at temperatures ranging from 20 to 750°C. In addition, pruning long short-term memory (P-LSTM) networks to create a predictive model for the stress–strain relationship of NWAC and LWAC is also utilized. Concrete mixture designs containing ordinary Portland cement, silica fume, and lightweight expanded clay aggregate, were first optimized to reduce the number of experiments using the response surface method. Subsequently, 30 mixture designs were fabricated and subjected to compression tests, following exposure to varying temperatures that ranged from 20 to 750°C, to evaluate their stress–strain relationship and determine associated mechanical properties. Experimental results were then utilized to develop a P-LSTM model used to forecast the stress–strain relationship of concrete at varying temperatures. The P-LSTM model developed in this study improved the prediction accuracy and stability beyond conventional LSTM model, which would be useful in the design and optimization of NWAC and LWAC structures. Additionally, the P-LSTM model has a lower computational cost and less likelihood of over-fitting as compared to typical LSTM networks. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
11. Lower Limb Motion Recognition Based on sEMG and CNN-TL Fusion Model.
- Author
-
Zhou, Zhiwei, Tao, Qing, Su, Na, Liu, Jingxuan, Chen, Qingzheng, and Li, Bowen
- Abstract
To enhance the classification accuracy of lower limb movements, a fusion recognition model integrating a surface electromyography (sEMG)-based convolutional neural network, transformer encoder, and long short-term memory network (CNN-Transformer-LSTM, CNN-TL) was proposed in this study. By combining these advanced techniques, significant improvements in movement classification were achieved. Firstly, sEMG data were collected from 20 subjects as they performed four distinct gait movements: walking upstairs, walking downstairs, walking on a level surface, and squatting. Subsequently, the gathered sEMG data underwent preprocessing, with features extracted from both the time domain and frequency domain. These features were then used as inputs for the machine learning recognition model. Finally, based on the preprocessed sEMG data, the CNN-TL lower limb action recognition model was constructed. The performance of CNN-TL was then compared with that of the CNN, LSTM, and SVM models. The results demonstrated that the accuracy of the CNN-TL model in lower limb action recognition was 3.76%, 5.92%, and 14.92% higher than that of the CNN-LSTM, CNN, and SVM models, respectively, thereby proving its superior classification performance. An effective scheme for improving lower limb motor function in rehabilitation and assistance devices was thus provided. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
12. Enhanced Short-Term Load Forecasting: Error-Weighted and Hybrid Model Approach.
- Author
-
Yu, Huiqun, Sun, Haoyi, Li, Yueze, Xu, Chunmei, and Du, Chenkun
- Subjects
- *
GREY relational analysis , *PRINCIPAL components analysis , *LOAD forecasting (Electric power systems) , *SEARCH algorithms , *PREDICTION models , *FORECASTING - Abstract
To tackle the challenges of high variability and low accuracy in short-term electricity load forecasting, this study introduces an enhanced prediction model that addresses overfitting issues by integrating an error-optimal weighting approach with an improved ensemble forecasting framework. The model employs a hybrid algorithm combining grey relational analysis and radial kernel principal component analysis to preprocess the multi-dimensional input data. It then leverages an ensemble of an optimized deep bidirectional gated recurrent unit (BiGRU), an enhanced long short-term memory (LSTM) network, and an advanced temporal convolutional neural network (TCN) to generate predictions. These predictions are refined using an error-optimal weighting scheme to yield the final forecasts. Furthermore, a Bayesian-optimized Bagging and Extreme Gradient Boosting (XGBoost) ensemble model is applied to minimize prediction errors. Comparative analysis with existing forecasting models demonstrates superior performance, with an average absolute percentage error (MAPE) of 1.05% and a coefficient of determination (R2) of 0.9878. These results not only validate the efficacy of our proposed strategy, but also highlight its potential to enhance the precision of short-term load forecasting, thereby contributing to the stability of power systems and supporting societal production needs. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
13. A New Algorithm for Predicting Dam Deformation Using Grey Wolf-Optimized Variational Mode Long Short-Term Neural Network.
- Author
-
Sun, Xiwen, Lu, Tieding, Hu, Shunqiang, Wang, Haicheng, Wang, Ziyu, He, Xiaoxing, Ding, Hongqiang, and Zhang, Yuntao
- Subjects
- *
PREDICTION models , *TIME series analysis , *PROBLEM solving , *DAMS , *ENTROPY - Abstract
To solve the problems of difficult to model parameter selections, useful signal extraction and improper-signal decomposition in nonlinear, non-stationary dam displacement time series prediction methods, we propose a new predictive model for grey wolf optimization and variational mode decomposition and long short-term memory (GVLSTM). Firstly, we used the grey wolf optimization (GWO) algorithm to optimize the parameters of variable mode decomposition (VMD), obtaining the optimal parameter combination. Secondly, we used multiscale permutation entropy (MPE) as a standard to select signal screening, determining and recon-structing the effective modal components. Finally, the long short-term memory neural network (LSTM) was used to learn the dam deformation characteristics. The result shows that the GVLSTM model can effectively reduce the estimation deviation of the prediction model. Compared with VMDGRU and VMDANN, the average RMSE and MAE value of each station is increased by 19.11%~28.58% and 27.66%~29.63%, respectively. We used determination (R2) coefficient to judge the performance of the prediction model, and the value of R2 was 0.95~0.97, indicating that our method has good performance in predicting dam deformation. The proposed method has outstanding advantages of high accuracy, reliability, and stability for dam deformation prediction. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
14. Comparing dimensionality reduction techniques for visual analysis of the LSTM hidden activity on multi-dimensional time series modeling.
- Author
-
Ji, Lianen, Qiu, Shirong, Xu, Zhi, Liu, Yue, and Yang, Guang
- Subjects
- *
MULTIDIMENSIONAL scaling , *VISUAL analytics , *PRINCIPAL components analysis , *TIME series analysis , *DATA modeling - Abstract
Long short-term memory (LSTM) network is widely applied to multi-dimensional time series modeling to solve many real-world problems, and visual analytics plays a crucial role in improving its interpretability. To understand the high-dimensional activations in the hidden layer of the model, the application of dimensionality reduction (DR) techniques is essential. However, the diversity of DR techniques dramatically increases the difficulty of selecting one among them. In this paper, aiming at the applicability of DR techniques for visual analysis of LSTM hidden activity on multi-dimensional time series modeling, we select four representative DR techniques as the comparison objects, including principal component analysis (PCA), multi-dimensional scaling (MDS), t-distributed stochastic neighbor embedding (t-SNE), and uniform manifold approximation and projection (UMAP). The original continuous modeling data and the symbolically processed discrete data are used as knowledge of model learning, which are associated with LSTM hidden layer activity, and the ability of DR techniques to maintain high-dimensional information of the hidden layer activation is compared. According to the model structure of LSTM and the characteristics of modeling data, the controlled experiments were carried out in five typical tasks, namely the quality evaluation of DR, the abstract representation of high and low hidden layers, the association analysis between model and output variable, the importance analysis of input features and the exploration of temporal regularity. Through the complete experimental process and detailed result analysis, we distilled a systematic guidance for analysts to select appropriate and effective DR techniques for visual analytics of LSTM. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
15. Adaptive weighted feature fusion for multiscale atrous convolution‐based 1DCNN with dilated LSTM‐aided fake news detection using regional language text information.
- Author
-
Rathinapriya, V and Kalaivani, J.
- Subjects
- *
LANGUAGE models , *CONVOLUTIONAL neural networks , *LONG short-term memory , *OPTIMIZATION algorithms , *FAKE news - Abstract
The people in the world rely on social media for gathering news, and it is mainly because of the development of technology. The approaches employed in natural language processing are still deficient in judgement factors, and these techniques frequently rely upon political or social circumstances. Numerous low‐level communities in the area are curious after experiencing the negative effects caused by the spread of false information in different sectors. Low‐resource languages are still distracted, because these techniques are extensively employed in the English language. This work aims to provide an analysis of regional language fake news and develop a referral system with advanced techniques to identify fake news in Hindi and Tamil. This proposed model includes (a) Regional Language Text Collection; (b) Text preprocessing; (c) Feature Extraction; (d) Weighted Stacked Feature Fusion; and (e) Fake News Detection. The text data is collected from the standard datasets. The collected text data is preprocessed and given into the feature extraction, which is done by using bidirectional encoder representations from transformers (BERT), transformer networks, and seq2seq network for extracting the three sets of language text features. These extracted feature sets are inserted into the weighted stacked feature fusion model, where the three sets of extracted features are integrated with the optimized weights that are acquired through the enhanced osprey optimization algorithm (EOOA). Finally, these resultant features are given to multi‐scale atrous convolution‐based one‐dimensional convolutional neural network with dilated long short‐term memory (MACNN‐DLSTM) for detecting the fake news. Throughout the result analysis, the experimentation is conducted based on the standard Tamil and Hindi datasets. Moreover, the developed model shows 92% for Hindi datasets and 96% for Tamil datasets which shows effective performance regarding accuracy measures. The experimental analysis is carried out by comparing with the conventional algorithms and detection techniques to showcase the efficiency of the developed regional language‐based fake news detection model. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
16. Detection of neurodegenerative diseases using hybrid MODWT and adaptive local binary pattern.
- Author
-
Prasanna, J., George, S. Thomas, and Subathra, M. S. P.
- Subjects
- *
DISCRETE wavelet transforms , *HUNTINGTON disease , *AMYOTROPHIC lateral sclerosis , *FEATURE extraction , *PARKINSON'S disease , *GAIT in humans - Abstract
Neurodegenerative diseases cause significant irregularities in walking patterns, impacting gait dynamics and rhythms analyzed through gait time series. Human gait analysis is a promising avenue for identifying unique walking patterns. Automated computer-aided techniques show potential in tracking pathological progression, particularly through non-invasive methods using football contact sensors. In this study, wavelet coefficients extracted via the maximal overlapped discrete wavelet transform from gait time series provide valuable insights into neurological deficiencies and deviations in gait. We employed various local binary patterns (LBPs), including inverse LBP, adaptive right-shifted LBP, and adaptive left-shifted LBP (ALS-LBP) on wavelet coefficients for classifying neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS), Huntington's disease (HD), Parkinson's disease (PD), and healthy control (HC). Histogram-oriented features were extracted using different binary pattern techniques on gait time series. The feature subset was classified using the long short-term memory classifier. The study achieved maximum accuracy across all experimental cases, analyzing signals from left, right, and both feet during stride, swing, and stance. This approach demonstrated 100% classification accuracy for tasks involving HC versus PD, ALS versus HD, and ALS versus PD. The proposed method could open avenues for early-stage diagnosis of neurodegenerative diseases. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
17. A deep learning LSTM-based approach for AMD classification using OCT images.
- Author
-
Hamid, Laila, Elnokrashy, Amgad, Abdelhay, Ehab H., and Abdelsalam, Mohamed M.
- Subjects
- *
CONVOLUTIONAL neural networks , *MACULAR degeneration , *OPTICAL coherence tomography , *VISION disorders , *DEEP learning - Abstract
Age-related macular degeneration (AMD) is an age-related, persistent, painless eye disease that impairs central vision. The central area (macula) of the retina, located at the back of the eye, sustains damage that is the cause of loss of vision. The early detection of AMD can increase the probability of treatment and prevent vision loss. The AMD can be classified into dry and wet AMD based on the absence of neovascularization. This study introduces a new methodology for the classification of AMD using optical coherence tomography (OCT) retinal images. The proposed methodology is based on three stages. The first stage is the data preparation stage for resizing and normalizing the used images. The second stage is the image processing stage for enhancing the image quality as contrast and resolution these enhancements have been checked by the weighted peak signal-to-noise ratio (WPSNR) methodology. The third stage is the deep feature extraction and classification stage, which consists of two sub-models. The first model is MobileNet V1 which has been used as a deep feature extractor. The second model is LSTM (long short-term memory), fed with deep features to classify the AMD stages. A multi-classification with six separate trials has been employed with the proposed methodology, and compared with other models like DenseNet201 and InceptionV3. The proposed model has been tested on a sample of benchmark data with 4005 grayscale images labeled into three classes. The proposed methodology has achieved an accuracy of 98.85%, a sensitivity of 99.09%, and a specificity of 99.1%. To ensure the effectiveness of the proposed methodology, a comparative analysis has been established with previous approaches in the related field, and the results demonstrated the superiority of the proposed system in AMD multi-classification. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
18. Travel Time Reliability of an urban bus route, a comparison of Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM) approaches.
- Author
-
Hajibagheri, F. and Mamdoohi, A. R.
- Subjects
- *
TRAVEL time (Traffic engineering) , *LONG short-term memory , *BOX-Jenkins forecasting , *TRANSPORTATION planning , *LOCATION data , *STANDARD deviations , *QUALITY of service - Abstract
Reliability is a key factor in public transportation systems, significantly influencing passenger satisfaction and their perceptions of service quality. Therefore, measures of Travel Time Reliability (TTR), which help quantify unexpected delays, are essential for effectively planning and managing travel times. This research aims to provide a prediction model for standard deviation of travel time, as an indicator of TTR, on an urban bus route using Automatic Vehicle Location data as well as evaluating and comparing the results of ARIMA and LSTM models. Our research contributes to the existing body Ofliterature by considering the effect of a wide range of predictor variables which were categorized by route characteristics, weather condition, congestion, and temporal feature, on bus TTR, which received less attention in previous studies. The results show that LSTM is an efficient and accurate TTR predictive model than ARIMA, with an accuracy of about 87%. In addition, the LSTM model exhibits lower mean absolute percentage error (40%), mean square error (52.08%), and root mean square error (39.70%) compared to the corresponding values in the ARIMA model. Also, the congestion, weather condition, and holidays are the key variables in increasing the accuracy of LSTM model, respectively. Our findings provide insights to facilitate the decision-making of managers and planners in public transportation planning sector to improve transit reliability and passengers' satisfaction level. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
19. Multicriteria client selection model using class topper optimization based optimal federated learning for healthcare informatics.
- Author
-
Narwaria, Mamta and Jaiswal, Shruti
- Subjects
- *
ARTIFICIAL neural networks , *CONVOLUTIONAL neural networks , *FEDERATED learning , *SHORT-term memory , *LONG short-term memory - Abstract
Quality of life (QoL) of patients has grown as a result of thoughtful medical care systems where many stakeholders remotely review records. Data privacy is highly at risk due to open communication channels, which also has an impact on how models are trained using centralized servers' acquired data. An emerging idea called federated learning (FL) provides a workable remedy to this problem. There hasn't been a comprehensive or in-depth study of FL in the field of health informatics (HI), in contrast to previous studies that mainly focused on the role of FL in diverse applications. In this proposed approach, a Class Topper Optimization (CTO) based federated learning approach is developed. Clinical data's uploaded by clients are taken as input for this proposed work. Stratified sampling is employed to select clients according to their metadata, preventing contacts with clients that aren't relevant. In this paper, clients are selected based on the CTO approach utilizing a variety of criteria's. The server then receives the newly created parameters from each selected clients, which is then utilized for the training process of the local model. Two different algorithms named as Artificial Neural Network (ANN) and Convolutional Neural Network (CNN) are utilized as a local model to train the homogeneous client. The global model is further improved periodically by utilizing the updates from the locally trained instances. Long Short Term Memory (LSTM) is employed as a global model here. The proposed approach achieves 93% accuracy and 92% precision. Thus, the proposed optimization based client selection approach is the best choice for federated learning. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
20. Multi-unit Wind Power Prediction Based on Long Short-term Memory and Particle Swarm Optimization.
- Author
-
WU Zhenlong, MO Yipeng, WANG Ronghua, FAN Xinyu, LIU Yanhong, and GUO Xiaolian
- Abstract
At present, the manual adjustment of hyper-parameter for current wind power prediction model was slow and unreliability. In order to achieve the prediction effect, the model used in wind power prediction needs to select the appropriate hyper-parameters for the model. Based on this, in this study, a multi-unit wind power prediction model was proposed based on long short-term memory (LSTM). Firstly, the Spearman correlation method was used to quantitative analysis. Secondly, the principal component analysis (PCA) was used to reduce the dimension of the input features as well as extract the key information. In addition, considering the difficulty of choosing parameters for LSTM, in this study, particle swarm optimization (PSO) algorithm was used to optimize the number of hidden layer neurons in each layer of LSTM. For the problem of wind power prediction of multiple units, in this study, a single wind turbine was used to find the most excellent model in a single unit, and applied the prediction model to multi-unit prediction. Experiments showed that compared with other models, the root mean square error of the proposed method was reduced by 11. 8%, and the mean absolute error was reduced by 5. 03% [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
21. A LONG SHORT TERM MEMORY MODEL FOR CHARACTER-BASED ANALYSIS OF DNS TUNNELING DETECTION.
- Author
-
TAYYEH, HUDA KADHIM and AHMED AL-JUMAILI, AHMED SABAH
- Subjects
SHORT-term memory ,LONG short-term memory ,INFORMATION technology security ,MACHINE learning ,TUNNEL design & construction - Abstract
DNS tunneling is the attempt to create a hidden tunnel through a domain name service. Such a tunnel would jeopardize the targeted network and open the door for illegal access, control, and data exfiltration. The information security research community showed the variety of techniques that have been proposed to detect the tunnel. The majority of these efforts were relying on machine learning techniques where features of tunneling are considered such as length of DNS query, size, and entropy of the query. However, an additional analysis of the lexical information of the DNS query has been depicted recently and showed remarkable performance. This paper aims to examine the role of Long Short Term Memory (LSTM) model in terms of DNS lexical analysis. Two benchmark datasets related to DNS have been used. In addition, a character mapping mechanism has been used to replace every possible character with an integer number. Consequentially, the mapped representation has been fed into an LSTM model for DNS tunneling detection. Results showed that the proposed method was able to obtain a weighted average F1-score of 98% for both datasets respectively. Such results are competitive in the context of the state of the art and demonstrate the efficacy of the lexical analysis within the DNS tunneling detection task. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
22. CHARACTER-LEVEL EMBEDDING USING FASTTEXT AND LSTM FOR BIOMEDICAL NAMED ENTITY RECOGNITION.
- Author
-
AHMED AL-JUMAILI, AHMED SABAH and TAYYEH, HUDA KADHIM
- Subjects
SHORT-term memory ,LONG short-term memory ,RESEARCH personnel - Abstract
Extracting biomedical entities has caught many researchers' attention in which the recent technique of word embedding is employed for such a task. Yet, the traditional word embedding architectures of Word2vec or Glove are still suffering from the 'out-of-vocabulary' (OOV) problem. This problem occurs when an unseen term might be encountered during the testing which leads to absence of embedding vector. Hence, this study aims to propose a character-level embedding through FastText architecture. In fact, handling the character-level seems a promising solution for the OOV problem. To this end, the proposed FastText architecture has been used to generate embedding vectors for the possible N-gram combinations of each word. Consequentially, these vectors have been fed to a Long Short Term Memory (LSTM) architecture for classifying the words into its biomedical classes. Using two benchmark datasets of BioCreative-II and NCBI, the proposed method was able to produce an f-measure of 0.912 and 0.918 respectively. Comparing these results with the baseline studies demonstrates the superiority of the proposed character-level embedding of FastText in terms of Biomedical Named Entity Recognition (BNER) task. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
23. DESIGN AND APPLICATION OF PARAMETER SELF-TUNING REGULATOR FOR DC MOTOR BASED ON NEURAL NETWORK.
- Author
-
XIAODONG YANG, WEIJING GE, and YULIN WANG
- Subjects
ARTIFICIAL neural networks ,SELF-tuning controllers ,ADAPTIVE control systems ,INSTRUCTIONAL systems ,MACHINE learning - Abstract
This study introduces a cutting-edge approach to regulating DC motors, featuring a unique combination of Artificial Neural Networks (ANN) and Long Short-Term Memory (LSTM) networks. This innovative system capitalizes on the adaptive learning capabilities of ANNs to dynamically fine-tune the control parameters of DC motors. This adaptability ensures optimal motor performance across diverse operational conditions, addressing the challenges posed by fluctuating loads and varying speed requirements. The integration of LSTM networks into this framework adds a layer of predictive functionality, allowing the system to anticipate future motor states. Such foresight enables the regulator to make proactive adjustments, significantly enhancing its responsiveness to changes in operational demands. The dual application of ANN's adaptive control mechanisms and LSTM's predictive capabilities is particularly effective in overcoming the non-linearity and variability that are typical challenges in DC motor control. This synergy ensures that the motor operates efficiently, stably, and with a quick response time, even under varying and unpredictable conditions. The practical application of this advanced regulator in real-world scenarios has shown marked improvements in motor performance. These enhancements are evident in the increased efficiency, stability, and responsiveness of the motors, making them more suitable for a wide range of industrial applications. This study marks a notable progression in the field of DC motor control technology. By integrating advanced machine learning techniques, it offers a solution that is not only more efficient and reliable but also adaptable to the evolving demands of industrial environments. The innovative combination of ANN and LSTM networks in this regulator design paves the way for smarter, more responsive, and efficient motor control systems, potentially transforming how motors are managed in various industrial applications. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
24. An Improved Toxic Speech Detection on Multimodal Scam Confrontation Data Using LSTM-Based Deep Learning.
- Author
-
Gumelar, Agustinus Bimo, Yuniarno, Eko Mulyanto, Nugroho, Arif, Adi, Derry Pramono, Sugiarto, Indar, and Purnomo, Mauridhi Hery
- Subjects
SOCIAL media ,INTERNET fraud ,SPEECH ,INDONESIAN language ,NATIVE language - Abstract
Toxic speech has gained substantial attention, focusing on its detrimental effects and prevalence across online platforms. This phenomenon often exhibits discernible patterns in pronunciation analogous to emotions such as happiness or anger. It has been relatively underexplored in prior studies, which predominantly addressed offensive language, hate speech, and sarcasm without considering their emotional properties. Social media platforms have emerged as spaces where individuals share personal encounters with toxic speech that impacts on their well-being. To address this challenge, our study introduces a novel approach that combines speech and text data within a Long Short-Term Memory (LSTM) framework. Unlike existing methods that primarily focus on text analysis, our approach uniquely integrates both speech and text, thereby enhancing the model's ability to accurately detect toxic content. This multimodal data strategy is such an innovative step forward that it provides a more comprehensive solution to the problem of toxic speech detection. Our collected dataset comprises two-way conversations from online fraud reports and confrontations related to loan scams uploaded on YouTube, conducted in the Indonesian language. The absence of subtitles can emerge any ambiguity of homonyms, so it is required to transcribe the audio content to text. To do this, we used native speakers to make sure the transcription was correct in the Indonesian language of the toxic context. In addition, speech features, such as pitch, intensity, and speaking rate, were utilized alongside text features, including Bag-of-Words (BoW) and Term Frequency-Inverse Document Frequency (TF-IDF). As a result, validation through F1-score measurement yielded 92.73% for text data and 89.09% for speech data. Our proposed approach provided a substantial improvement of approximately 12%-30% compared to the previous LSTM models. The performance comparison results confirmed that our proposed approach can enhance the accuracy of toxic speech detection. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
25. The Deep Learning Based Epileptic Seizure Detection Using 2-layer Convolutional Network with Long Short-Term Memory.
- Author
-
Vaithilingam, Sonia Devi and Regulagedda, Pallavi
- Subjects
CONVOLUTIONAL neural networks ,SEIZURES (Medicine) ,EPILEPSY ,CENTRAL nervous system ,NEUROLOGICAL disorders ,DEEP learning - Abstract
Epilepsy is a pervasive chronic neurological disorder characterized through irregular electrical discharges in the brain which causes seizures. Epilepsy seizure is a disorder that affects the brain cells with an influence on an effectiveness of central nervous system. Electroencephalography (EEG) is a majorly utilized method for epileptic seizure detection and diagnosis. In this research, Deep Learning (DL) methods of 2-layer Convolutional Neural Network (CNN) with Long Short-Term Memory (LSTM) are proposed for an automatic detection and diagnosis of an epileptic seizure. In the pre-processing phase, a Butterworth filter method of order 2 is used to remove noise in the EEG signal. The 2-layer CNN is used for the process of feature extraction. In 2-layer LSTM, one layer is utilized to perform short-term dependencies, while another layer is utilized to perform long term dependencies. In the end, the proposed method classifies seizures into epileptic and non-epileptic. The results demonsrates that the proposed method delivers performance metrics of better accuracy of 99.90% and sensitivity of 90.06% using CHB-MIT and Bonn datasets which contains EEG signals as compared to the existing methods like CNN and Epilepsy-Net. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
26. Nonlinear Dynamic Weight-Salp Swarm Algorithm and Long Short-Term Memory with Gaussian Error Linear Units for Network Intrusion Detection System.
- Author
-
Satyanarayana, Darala and Saikiran, Ellambotla
- Subjects
COMPUTER network security ,FEATURE selection ,DEEP learning ,INTRUSION detection systems (Computer security) ,ALGORITHMS ,CYBERTERRORISM - Abstract
Network Intrusion Detection Systems (NIDS) are essential for defending cyberattacks, offering vital protection for network security. Recently, there has been a significant growth in the usage of Deep Learning (DL) based algorithms for intrusion detection within the network security. Intrusion detection is a challenging task due to irrelevant or inappropriate features being used for the classification process. In this research, the Nonlinear Dynamic Weight -- Salp Swarm Algorithm (NDW-SSA) based feature selection method is employed to select the relevant or appropriate features for classification. The dynamic weight is included in the traditional SSA which enhances the performance of the SSA to select relevant features for classification. Then, the Long Short-Term Memory with Gaussian Error Linear Units (LSTM with GELU) method is developed for the classification of intrusion types. The GELU activation function is utilized during a training process of LSTM which reduces gradient vanishing issue and stabilizes the training process. The NDW-SSA and LSTM with the GELU method obtains 96.78% accuracy on the NSL-KDD dataset, 98.78% accuracy on the UNSW-NB15 dataset, and 98.77% accuracy on the CIC-IDS 2017 dataset, which is superior when compared to Local Search -- Pigeon Inspired Optimization (LS-PIO). [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
27. Smart bridge bearing monitoring: Predicting seismic responses with a multi‐head attention‐based CNN‐LSTM network.
- Author
-
Yazdanpanah, Omid, Chang, Minwoo, Park, Minseok, and Mangalathu, Sujith
- Subjects
SHORT-term memory ,LONG short-term memory ,AXIAL loads ,RUBBER bearings ,LATERAL loads ,SEISMIC response - Abstract
This paper introduces a novel method to spontaneously predict displacement time histories and hysteresis curves of bridge lead rubber bearings under seismic loads and axial forces. The method leverages a stacked convolutional‐bidirectional Cuda Long Short Term Memory network, enhanced with multi‐head attention, skip connections, exponential learning rate scheduler, and a hybrid activation function to improve performance. The framework utilizes the functional application programming interface provided by the Python Keras library to build a model that takes input features such as horizontal and vertical ground accelerations, actuator loads in both lateral and vertical directions, and the superstructure mass. The effectiveness of the deep learning model is evaluated using a considerable experimental dataset of 53 real‐time hybrid simulations, spanning various earthquake intensities and superstructure masses (Chi‐Chi: 15 scenarios, El Centro: 15 scenarios, Kobe: 13 scenarios, and Northridge: 10 scenarios). Initially, Northridge earthquake data serves as unseen data, while the rest is used for training and validation. In a subsequent trial, the unseen data is centered on Kobe earthquake scenarios. By employing a hybrid loss function merging mean square and mean absolute errors, the model exhibits a substantial correlation of over 83% between predicted displacement time series and empirical measurements for the unseen data. In summary, the proposed model offers miscellaneous benefits, including time and cost savings in experimental efforts by decreasing the need for additional tests. It further delivers a swift and precise insight into the bridge bearing performance and its energy dissipation, facilitating timely and accurate bridge design in different scenarios for engineers. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
28. Enhanced attention-driven hybrid deep learning with harris hawks optimizer for apple mechanical damage detection.
- Author
-
Ma, Ling, Wu, Xincan, Zhu, Ting, Huang, Yingxinxin, Chen, Xinnan, Ning, Jingyuan, Sun, Yuqi, and Hui, Guohua
- Subjects
METAHEURISTIC algorithms ,OPTIMIZATION algorithms ,DEEP learning ,NONDESTRUCTIVE testing ,NEAR infrared spectroscopy - Abstract
This study addresses the challenges of high costs and lengthy detection times associated with non-destructive testing of mechanical damage in apples. A novel approach combining deep learning and the Harris hawks optimizer (HHO) is proposed to tackle this. The study employs near-infrared relaxation spectroscopy to analyze apples' spectral characteristics in different conditions. These spectral data are then processed by a residual network (ResNet) to extract relevant features. The extracted features are subsequently fed into a fusion model comprising long short-term memory (LSTM) and an Attention mechanism, with the model's output determined by the Softmax function. The HHO is utilized to optimize parameter combinations for the search models, and its performance is compared against the gray wolf optimization algorithm whale optimization algorithm (WOA), and dwarf mongoose optimization algorithm. Moreover, the study introduces the Multiple Measurement Classification Recognition (MMCR) method to enhance accuracy. Comparative analyses demonstrate that the HHO-ResNet-LSTM (Attention)-MMCR model effectively captures intricate nonlinear relationships, resulting in an impressive accuracy increase to 98%. This innovative model offers a promising avenue for non-destructive fruit inspection, contributing to the advancement of inspection methodologies. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
29. CNN--LSTM Hybrid Deep Learning Model for Detergent Products Demand Forecasting: A Case Study.
- Author
-
Ghazouani, Imen, Masmoudi, Imen, Mejri, Imen, and Layeb, Safa Bhar
- Subjects
ARTIFICIAL neural networks ,CONVOLUTIONAL neural networks ,DEMAND forecasting ,COVID-19 pandemic ,PRODUCTION scheduling ,INVENTORY control - Abstract
An accurate forecast of current and future demand is an essential initial step for almost all the facets of supply chain optimization, including inventory strategy, production scheduling, distribution management, and marketing policies. Simply put, a more accurate demand prediction can lead to a more optimized supply chain process, allowing for better inventory control and higher customer satisfaction. Classical demand predictions are based principally on qualitative approaches relying on data from experts' opinions; quantitative forecasts based on historical data through statistical and artificial neural network models or a mix of qualitative and quantitative techniques that is also widely used and has shown good performances. Detergents and cleaning products demand is extremely volatile and has undergone substantial change, especially during the COVID-19 health crisis. In this paper, we present a hybrid Neural Network approach for accurate demand forecasts of the detergent manufacturing industry. It mainly consists of the combination of Long Short-Term Memory (LSTM) with Convolution Neural Network (CNN) based approaches. We performed a series of experiments on real data sets and assessed the performance of the proposed CNN-LSTM hybrid model. Numerical results showed that the combination of LSTM layers with complementary CNN layers provides more accurate results than other state-of-the-art forecasting models. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
30. Blind image quality assessment using Beltrami filter-based contrast features (BF-bCF) & LSTM network.
- Author
-
Gabhane, Yogita, Jain, Tapan Kumar, and Kamble, Vipin
- Subjects
- *
PRINCIPAL components analysis , *FEATURE extraction , *GRAYSCALE model , *DOGS - Abstract
Advanced networks excel in Blind Image Quality Assessment (BIQA), however accurate estimation of quality scores remains challenging due to unrevealed features extricated under various distortions. This article presents a Beltrami filter-based Contrast feature and long short-term memory (BF-bCF & LSTM) framework for feature extraction, suitable for both authentic and synthetic image distortions. The framework comprises four modules: an edge preservation and enhancement module, a distortion-conscious module, a weight module, and a quality-predicting network. The edge enhancement module assists the distortion-conscious module by enhancing the overall quality of the input grayscale image. The distortion-conscious module uses a modified Difference of Gaussian (DoG) measure to mitigate the distortion and assign patch weights in the filtered image. The proposed BIQA framework assimilating LSTM effectively extricates significant components using principal component analysis (PCA) and predicts nearer scores. Experiments on synthetic and authentic datasets showed superior performance relating to SROCC and PLCC above 0.9. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
31. AUTOMATED GESTURE RECOGNITION USING APPLIED LINGUISTICS WITH DATA-DRIVEN DEEP LEARNING FOR ARABIC SPEECH TRANSLATION.
- Author
-
ALAHMARI, SAAD, AL-ONAZI, BADRIYYA B., ALJOHANI, NOUF J., ALZAHRANI, KHADIJA ABDULLAH, ALOTAIBI, FAIZ ABDULLAH, ALMANEA, MANAR, ALNFIAI, MRIM M., and MAHGOUB, HANY
- Subjects
- *
SIGN language , *MACHINE learning , *COMPUTER vision , *FACIAL expression , *SPEECH perception , *DEEP learning - Abstract
Gesture recognition for Arabic speech translation includes developing advanced technologies that correctly translate body and hand movements corresponding to Arabic sign language (ArSL) into spoken Arabic. This leverages machine learning and computer vision techniques in complex systems simulation platforms to scrutinize the gestures utilized in ArSL, detecting mild differences in facial expressions, hand shapes, and movements. Sign Language Recognition (SLR) is paramount in assisting communication for the Deaf and Hard-of-Hearing communities. It includes using vision-based methods and Surface Electromyography (sEMG) signals. The sEMG signal is crucial for recognizing hand gestures and capturing muscular activities in sign language. Researchers have comprehensively shown the capability of EMG signals to approach specific details, mainly in classifying hand gestures. This progression is a stimulating feature in extracting the interpretation and recognition of sign languages and investigating the phonology of signed language. Leveraging machine learning algorithms and signal processing techniques in complex systems simulation platforms, researchers aim to extract relevant traits from the sEMG signals that correspond to different ArSL gestures. This study introduces an Enhanced Dwarf Mongoose Algorithm with a Deep Learning-Driven Arabic Sign Language Detection (EDMODL-ASLD) technique on sEMG data. In the initial phase, the presented EDMODL-ASLD model is subjected to data preprocessing to change the input sEMG data into an attuned format. In the next stage, feature extraction with fractal theories is used to gather relevant and nonredundant data from the EMG window to construct a feature vector. In this study, the absolute envelope (AE), energy (E), root-mean square (RMS), standard deviation (STD), and mean absolute value (MAV) are the five time-domain extracted features for the EMG window observation. Meanwhile, the dilated convolutional long short-term memory (ConvLSTM) technique is used to identify distinct sign languages. To improve the results of the dilated ConvLSTM model, the hyperparameter selection process is executed using the EDMO model. To illustrate the significance of the EDMODL-ASLD technique, a brief experimental validation is made on the Arabic SLR database. The experimental validation of the EDMODL-ASLD technique portrayed a superior accuracy value of 96.47% over recent DL approaches. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
32. Sentiment analysis deep learning model based on a novel hybrid embedding method.
- Author
-
Ouni, Chafika, Benmohamed, Emna, and Ltifi, Hela
- Abstract
(WE) are crucial for capturing the meanings of words, offering continuous vector representations that encode both semantic and syntactic information. In this paper, we present a novel approach called WordFast, which combines the strengths of FastText and Word2Vec through a linear combination method. The WordFast approach aims to enhance the performance of WE, particularly in the context of sentiment analysis (SA). SA has become a prominent area of research in Natural Language Processing (NLP), especially when it comes to analyzing user opinions on digital platforms. Our proposed (SA) deep model is based on the WordFast method and incorporates two variations of Recurrent Neural Network (RNN) architectures. This model is tested using two datasets: IMDB reviews and Amazon reviews.The outcomes produced by the WordFast method are classified using Gated Recurrent Unit (GRU) and Long Short-Term Memory (LSTM) models.Our experiments reveal a significant improvement in accuracy when analyzing real IMDB, achieving 88.75/% and 89.54%, as well as real Amazon reviews, with accuracies of 94.69% and 94.89%. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
33. BiLSTM-based selection and prediction of optimal polling systems for multiple server numbers.
- Author
-
Yang, Zhijun, Huang, Wenjie, and Ding, Hongwei
- Subjects
ARTIFICIAL neural networks ,SHORT-term memory ,LONG short-term memory ,5G networks ,MATHEMATICAL analysis - Abstract
Facing the demands of different service scenarios and the large number of base stations deployed in the context of the 5G era, network slicing was introduced to support on-demand services for specific service scenarios. The base stations in this context, are utilized to provide services within different slices in each service scenario. Due to the complexity of selection of the most appropriate base stations, a multi-server polling system for 5G network slicing is proposed in this work, with a method of predicting the optimal number of base stations to be selected. In order of solve the problem of difficult mathematical analysis, low service efficiency, high delay and waste of resources due to unlimited number of base stations in the network, the selection method makes use of a Bi-directional Long Short Term Memory (BiLSTM) neural network. First, the experiment was scaled up and multidimensional variables were varied to obtain data on the average queue length and average delay of the system; Next, a neural network model is constructed for performance prediction; Finally, the optimal number of base stations is selected and a posture prediction is made for that number of base stations. The experimental results show that the system performance is best when the number of base stations is 3, which saves resources and reduces the difficulty of mathematical derivation and analysis. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
34. Artificial Neural Networks for Energy Demand Prediction in an Economic MPC‐Based Energy Management System.
- Author
-
Alarcón, Rodrigo G., Alarcón, Martín A., González, Alejandro H., and Ferramosca, Antonio
- Subjects
- *
ARTIFICIAL neural networks , *ECONOMIC forecasting , *ECONOMIC models , *ENERGY management , *ARTIFICIAL intelligence , *RECURRENT neural networks - Abstract
ABSTRACT Microgrids are a development trend and have attracted a lot of attention worldwide. The control system plays a crucial role in implementing these systems and, due to their complexity, artificial intelligence techniques represent some enabling technologies for their future development and success. In this paper, we propose a novel formulation of an economic model predictive control (economic MPC) applied to a microgrid designed for a faculty building with the inclusion of a predictive model to deal with the energy demand disturbance using a recurrent neural network of the long short‐term memory (RNN‐LSTM). First, we develop a framework to identify an RNN‐LSTM using historical data registered by a smart three‐phase power quality analyzer to provide feedforward power demand predictions. Next, we present an economic MPC formulation that includes the prediction model for the disturbance within the optimization problem to be solved by the MPC strategy. We carried out simulations with different scenarios of energy consumption, available resources, and simulation times to highlight the results obtained and analyze the performance of the energy management system. In all cases, we observed the correct operation of the proposed control scheme, complying at all times with the objectives and operational restrictions imposed on the system. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
35. LSTM-based framework for predicting point defect percentage in semiconductor materials using simulated XRD patterns.
- Author
-
Motamedi, Mehran, Shidpour, Reza, and Ezoji, Mehdi
- Subjects
- *
SEMICONDUCTOR materials , *X-ray diffraction , *POINT defects , *SPHALERITE , *CRYSTAL defects - Abstract
In this paper, we present a machine learning-based approach that leverages Long Short-Term Memory (LSTM) networks combined with a sliding window technique for feature extraction, aimed at accurately predicting point defect percentages in semiconductor materials based on simulated X-ray Diffraction (XRD) data. The model was initially trained on silicon-simulated XRD data with defect percentages ranging from 1 to 5%, enabling it to predict defect percentages from 0 to 10% in silicon and other semiconductor materials, including AlAs, CdS, GaAs, Ge, and ZnS. Through extensive experimentation, we explored different sequence lengths and LSTM units, identifying the optimal configuration as a sequence length of 3501 and 4500 units, which yielded the best results. The model's mean absolute error at 4500 units was 0.021, the lowest among the LSTM configurations tested. The sliding window technique plays a crucial role in capturing temporal dependencies within the XRD data, allowing the model to generalize to other semiconductor materials. Additionally, we observed that increasing defect percentages consistently led to a rise in background intensity. We further examined the relationship between crystal structure and defect precentage predictions, uncovering consistent trends for materials with Diamond Cubic and Zinc Blende structures. This LSTM-based method offers a novel approach to predicting defect percentages using simulated XRD patterns of materials. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
36. Implementing PSO-LSTM-GRU Hybrid Neural Networks for Enhanced Control and Energy Efficiency of Excavator Cylinder Displacement.
- Author
-
Nguyen, Van-Hien, Do, Tri Cuong, and Ahn, Kyoung-Kwan
- Subjects
- *
PARTICLE swarm optimization , *PEARSON correlation (Statistics) , *ENERGY management , *PID controllers , *PREDICTION models - Abstract
In recent years, increasing attention has been given to reducing energy consumption in hydraulic excavators, resulting in extensive research in this field. One promising solution has been the integration of hydrostatic transmission (HST) and hydraulic pump/motor (HPM) configurations in parallel systems. However, these systems face challenges such as noise, throttling losses, and leakage, which can negatively impact both tracking accuracy and energy efficiency. To address these issues, this paper introduces an intelligent real-time prediction framework for system positioning, incorporating particle swarm optimization (PSO), long short-term memory (LSTM), a gated recurrent unit (GRU), and proportional–integral–derivative (PID) control. The process begins by analyzing real-time system data using Pearson correlation to identify hyperparameters with medium to strong correlations to the positioning parameters. These selected hyperparameters are then used as inputs for forecasting models. Independent LSTM and GRU models are subsequently developed to predict the system's position, with PSO optimizing four key hyperparameters of these models. In the final stage, the PSO-optimized LSTM-GRU models are employed to perform real-time intelligent predictions of motion trajectories within the system. Simulation and experimental results show that the model achieves a prediction deviation of less than 3 mm, ensuring precise real-time predictions and providing reliable data for system operators. Compared to traditional PID and LSTM-GRU-PID controllers, the proposed controller demonstrated superior tracking accuracy while also reducing energy consumption, achieving energy savings of up to 10.89% and 2.82% in experimental tests, respectively. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
37. Performance Analysis and Prediction of 5G Round-Trip Time Based on the VMD-LSTM Method.
- Author
-
Zhu, Sanying, Zhou, Shutong, Wang, Liuquan, Zang, Chenxin, Liu, Yanqiang, and Liu, Qiang
- Subjects
- *
NETWORK performance , *TELECOMMUNICATION systems , *COMPUTER network protocols , *DATA transmission systems , *SHORT-term memory - Abstract
With the increasing level of industrial informatization, massive industrial data require real-time and high-fidelity wireless transmission. Although some industrial wireless network protocols have been designed over the last few decades, most of them have limited coverage and narrow bandwidth. They cannot always ensure the certainty of information transmission, making it especially difficult to meet the requirements of low latency in industrial manufacturing fields. The 5G technology is characterized by a high transmission rate and low latency; therefore, it has good prospects in industrial applications. To apply 5G technology to factory environments with low latency requirements for data transmission, in this study, we analyze the statistical performance of the round-trip time (RTT) in a 5G-R15 communication system. The results indicate that the average value of 5G RTT is about 11 ms, which is less than the 25 ms of WIA-FA. We then consider 5G RTT data as a group of time series, utilizing the augmented Dickey–Fuller (ADF) test method to analyze the stability of the RTT data. We conclude that the RTT data are non-stationary. Therefore, firstly, the original 5G RTT series are subjected to first-order differencing to obtain differential sequences with stronger stationarity. Then, a time series analysis-based variational mode decomposition–long short-term memory (VMD-LSTM) method is proposed to separately predict each differential sequence. Finally, the predicted results are subjected to inverse difference to obtain the predicted value of 5G RTT, and a predictive error of 4.481% indicates that the method performs better than LSTM and other methods. The prediction results could be used to evaluate network performance based on business requirements, reduce the impact of instruction packet loss, and improve the robustness of control algorithms. The proposed early warning accuracy metrics for control issues can also be used to indicate when to retrain the model and to indicate the setting of the control cycle. The field of industrial control, especially in the manufacturing industry, which requires low latency, will benefit from this analysis. It should be noted that the above analysis and prediction methods are also applicable to the R16 and R17 versions. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
38. Monitoring terrestrial water storage changes using GNSS vertical coordinate time series in Amazon River basin.
- Author
-
Liu, Yifu, Xu, Keke, Guo, Zengchang, Li, Sen, and Zhu, Yongzhen
- Subjects
- *
GLOBAL Positioning System , *LONG short-term memory , *PRINCIPAL components analysis , *DEFORMATION of surfaces , *GRAVITATIONAL fields , *WATER storage - Abstract
Aiming at the Terrestrial Water Storage(TWS) changes in the Amazon River basin, this article uses the coordinate time series data of the Global Navigation Satellite System (GNSS), adopts the Variational Mode Decomposition and Bidirectional Long and Short Term Memory(VMD-BiLSTM) method to extract the vertical crustal deformation series, and then adopts the Principal Component Analysis(PCA) method to invert the changes of terrestrial water storage in the Amazon Basin from July 15, 2012 to July 25, 2018. Then, the GNSS inversion results were compared with the equivalent water height retrieved from Gravity Recovery and Climate Experiment (GRACE) data. The results show that (1) the extraction method proposed in this article has better denoising effect than the traditional method; (2) the surface hydrological load deformation can be well calculated using GNSS coordinate vertical time series, and then the regional TWS changes can be inverted, which has a good consistency with the result of GRACE inversion of water storage, and has almost the same seasonal variation characteristics; (3) There is a strong correlation between TWS changes retrieved by GNSS based on surface deformation characteristics and water mass changes calculated by GRACE based on gravitational field changes, but GNSS satellite's all-weather measurement results in a finer time scale compared with GRACE inversion results. In summary, GNSS can be used as a supplementary technology for monitoring terrestrial water storage changes, and can complement the advantages of GRACE technology. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
39. Deep learning hybrid models with multivariate variational mode decomposition for estimating daily solar radiation.
- Author
-
Band, Shahab S., Qasem, Sultan Noman, Ameri, Rasoul, Pai, Hao-Ting, Gupta, Brij B., Mehdizadeh, Saeid, and Mosavi, Amir
- Subjects
STANDARD deviations ,SOLAR radiation ,ARTIFICIAL intelligence ,RENEWABLE energy sources ,MACHINE learning ,SOLAR energy - Abstract
Solar energy is one of the renewable and clean energy sources. Accurate solar radiation (SR) estimates are therefore needed in solar energy applications. Firstly, two deep learning models, including gated recurrent unit (GRU) and long short-term memory (LSTM), were developed in this study. Next, a data pre-processing technique named multivariate variational mode decomposition (MVMD) was used to construct the MVMD-GRU and MVMD-LSTM hybrid models. To better test the performance of proposed simple and hybrid models, four stations located in the Illinois State of the USA (i.e., Dixon Springs, Fairfield, Rend Lake, and Carbondale) were considered as the study sites. Whole the simple and hybrid models were established under two different strategies, i.e., local and external. In the local strategy, SR of each location was estimated using the minimum and maximum air temperatures from the same station. While, minimum and maximum air temperatures as well as SR data from the nearby station were utilized in external strategy to estimate SR time series of any target site. Root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R
2 ) metrics were used when evaluating the models performances. The overall results revealed that the proposed MVMD-GRU and MVMD-LSTM hybrid models illustrated better SR estimates compared to the simple GRU and LSTM in both the local and external strategies. The values of error metrics obtained for the superior hybrid models (i.e., MVMD-LSTM) during the testing period were as: RMSE = 2.532 MJ/m2 .day, MAE = 1.921 MJ/m2 .day, R2 = 0.916 at Dixon Springs; RMSE = 2.476 MJ/m2 .day, MAE = 1.878 MJ/m2 .day, R2 = 0.921 at Fairfield; RMSE = 2.359 MJ/m2 .day, MAE = 1.780 MJ/m2 .day, R2 = 0.924 at Rend Lake; RMSE = 2.576 MJ/m2 .day, MAE = 1.941 MJ/m2 .day, R2 = 0.914 at Carbondale. Therefore, the coupled models proposed in this study can be possibly recommended as suitable alternatives to the simple deep learning models with a reliable precision in estimating SR time series. [ABSTRACT FROM AUTHOR]- Published
- 2024
- Full Text
- View/download PDF
40. Classification of tomato seedling chilling injury based on chlorophyll fluorescence imaging and DBO-BiLSTM.
- Author
-
Zhenfen Dong, Jing Zhao, Wenwen Ji, Wei Wei, and Yuheng Men
- Subjects
FROST resistance of plants ,LONG short-term memory ,CHLOROPHYLL spectra ,IMAGE recognition (Computer vision) ,DUNG beetles ,DEEP learning - Abstract
Introduction: Tomatoes are sensitive to low temperatures during their growth process, and low temperatures are one of the main environmental limitations affecting plant growth and development in Northeast China. Chlorophyll fluorescence imaging technology is a powerful tool for evaluating the efficiency of plant photosynthesis, which can detect and reflect the effects that plants are subjected to during the low temperature stress stage, including early chilling injury. Methods: This article primarily utilizes the chlorophyll fluorescence image set of tomato seedlings, applying the dung beetle optimization (DBO) algorithm to enhance the deep learning bidirectional long short term memory (BiLSTM) model, thereby improving the accuracy of classification prediction for chilling injury in tomatoes. Firstly, the proportion of tomato chilling injury areas in chlorophyll fluorescence images was calculated using a threshold segmentation algorithm to classify tomato cold damage into four categories. Then, the features of each type of cold damage image were filtered using SRCC to extract the data with the highest correlation with cold damage. These data served as the training and testing sample set for the BiLSTM model. Finally, DBO algorithm was applied to enhance the deep learning BiLSTM model, and the DBO-BiLSTM model was proposed to improve the prediction performance of tomato seedling category labels. Results: The results showed that the DBO-BiLSTM model optimized by DBO achieved an accuracy, precision, recall, and F1 score with an average of over 95%. Discussion: Compared to the original BiLSTM model, these evaluation parameters improved by 9.09%, 7.02%, 9.16%, and 8.68%, respectively. When compared to the commonly used SVM classification model, the evaluation parameters showed an increase of 6.35%, 7.33%, 6.33%, and 6.5%, respectively. This study was expected to detect early chilling injury through chlorophyll fluorescence imaging, achieve automatic classification and labeling of cold damage data, and lay a research foundation for in-depth research on the cold damage resistance of plants themselves and exploring the application of deep learning classification methods in precision agriculture. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
41. Landslide Deformation Analysis and Prediction with a VMD-SA-LSTM Combined Model.
- Author
-
Wen, Chengzhi, Tian, Hongling, Zeng, Xiaoyan, Xia, Xin, Hu, Xiaobo, and Pang, Bo
- Subjects
GREY relational analysis ,FEATURE extraction ,LANDSLIDES ,PREDICTION models ,EXTRAPOLATION - Abstract
The evolution of landslides is influenced by the complex interplay of internal geological factors and external triggering factors, resulting in nonlinear dynamic changes. Although deep learning methods have demonstrated advantages in predicting multivariate landslide displacement, their performance is often constrained by the challenges of extracting intricate features from extended time-series data. To address this challenge, we propose a novel displacement prediction model that integrates Variational Mode Decomposition (VMD), Self-Attention (SA), and Long Short-Term Memory (LSTM) networks. The model first employs VMD to decompose cumulative landslide displacement into trend, periodic, and stochastic components, followed by an assessment of the correlation between these components and the triggering factors using grey relational analysis. Subsequently, the self-attention mechanism is incorporated into the LSTM model to enhance its ability to capture complex dependencies. Finally, each displacement component is fed into the SA-LSTM model for separate predictions, which are then reconstructed to obtain the cumulative displacement prediction. Using the Zhonghai Village tunnel entrance (ZVTE) landslide as a case study, we validated the model with displacement data from GPS point 105 and made predictions for GPS point 104 to evaluate the model's generalization capability. The results indicated that the RMSE and MAPE for SA-LSTM, LSTM, and TCN-LSTM at GPS point 105 were 0.3251 and 1.6785, 0.6248 and 2.9130, and 1.1777 and 5.5131, respectively. These findings demonstrate that SA-LSTM outperformed the other models in terms of complex feature extraction and accuracy. Furthermore, the RMSE and MAPE at GPS point 104 were 0.4232 and 1.0387, further corroborating the model's strong extrapolation capability and its effectiveness in landslide monitoring. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
42. A hybrid approach for early-identification of at-risk dropout students using LSTM-DNN networks.
- Author
-
El Aouifi, Houssam, El Hajji, Mohamed, and Es-Saady, Youssef
- Subjects
SCHOOL dropouts ,DEEP learning ,MACHINE learning ,PREDICTION models ,RANDOM forest algorithms - Abstract
Dropout refers to the phenomenon of students leaving school before completing their degree or program of study. Dropout is a major concern for educational institutions, as it affects not only the students themselves but also the institutions' reputation and funding. Dropout can occur for a variety of reasons, including academic, financial, personal, and social factors. Therefore, understanding the factors that contribute to dropout and developing effective strategies to prevent it is a critical challenge for educational institutions. In this study, we propose a hybrid deep learning model based on Long Short-Term Memory and Deep Neural Network algorithms for school dropout prediction. The proposed model was compared with previous works and several other machine learning algorithms, including Deep Neural Network (DNN), K-Nearest Neighbors (KNN), Naive Bayes (NB), Multi-Layer Perceptron (MLP), Decision Trees (DT), Support Vector Machine (SVM), and Random Forest (RF). The results showed that the proposed DNN-LSTM model outperforms the other models in terms of accuracy and efficiency. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
43. Sequence-to-Sequence Models and Their Evaluation for Spoken Language Normalization of Slovenian.
- Author
-
Sepesy Maučec, Mirjam, Verdonik, Darinka, and Donaj, Gregor
- Subjects
STATISTICAL hypothesis testing ,TRANSFORMER models ,SPEECH ,ORAL communication ,STATISTICAL models - Abstract
Sequence-to-sequence models have been applied to many challenging problems, including those in text and speech technologies. Normalization is one of them. It refers to transforming non-standard language forms into their standard counterparts. Non-standard language forms come from different written and spoken sources. This paper deals with one such source, namely speech from the less-resourced highly inflected Slovenian language. The paper explores speech corpora recently collected in public and private environments. We analyze the efficiencies of three sequence-to-sequence models for automatic normalization from literal transcriptions to standard forms. Experiments were performed using words, subwords, and characters as basic units for normalization. In the article, we demonstrate that the superiority of the approach is linked to the choice of the basic modeling unit. Statistical models prefer words, while neural network-based models prefer characters. The experimental results show that the best results are obtained with neural architectures based on characters. Long short-term memory and transformer architectures gave comparable results. We also present a novel analysis tool, which we use for in-depth error analysis of results obtained by character-based models. This analysis showed that systems with similar overall results can differ in the performance for different types of errors. Errors obtained with the transformer architecture are easier to correct in the post-editing process. This is an important insight, as creating speech corpora is a time-consuming and costly process. The analysis tool also incorporates two statistical significance tests: approximate randomization and bootstrap resampling. Both statistical tests confirm the improved results of neural network-based models compared to statistical ones. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
44. A Comprehensive Hybrid Deep Learning Approach for Accurate Status Predicting of Hydropower Units.
- Author
-
Ma, Liyong, Chen, Siqi, Wei, Dali, Zhang, Yanshuo, and Guo, Yinuo
- Subjects
CLEAN energy ,SUSTAINABILITY ,ELECTRIC power ,SHORT-term memory ,WATER power - Abstract
Hydropower units are integral to sustainable energy production, and their operational reliability hinges on accurate status prediction. This paper introduces an innovative hybrid deep learning model that synergistically integrates a Temporal Convolutional Network (TCN), a Residual Short-Term LSTM (REST-LSTM) network, a Gated Recurrent Unit (GRU) network, and the tuna swarm optimization (TSO) algorithm. The model was meticulously designed to capture and utilize temporal features inherent in time series data, thereby enhancing predictive performance. Specifically, the TCN effectively extracts critical temporal features, while the REST-LSTM, with its residual connections, improves the retention of short-term memory in sequence data. The parallel incorporation of GRU further refines temporal dynamics, ensuring comprehensive feature capture. The TSO algorithm was employed to optimize the model's parameters, leading to superior performance. The model's efficacy was empirically validated using three datasets—unit flow rate, guide vane opening, and maximum guide vane water temperature—sourced from the Huadian Electric Power Research Institute. The experimental results demonstrate that the proposed model significantly reduces both the maximum and average prediction errors, while also offering substantial improvements in forecasting accuracy compared with the existing methodologies. This research presents a robust framework for hydropower unit operation prediction, advancing the application of deep learning in the hydropower sector. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
45. Comparing the efficiency of recurrent neural networks to EMG-based continuous estimation of the elbow angle.
- Author
-
Davarinia, Fatemeh and Maleki, Ali
- Subjects
- *
FEEDFORWARD neural networks , *DELAY lines , *SHOULDER girdle , *LONG short-term memory , *ELBOW - Abstract
This study comprehensively assesses various recurrent neural networks (RNNs) for decoding the elbow angle from electromyogram (EMG) signals, a crucial aspect in myoelectric interfaces. EMG signals from the shoulder girdle and arm were recorded during goal-directed reaching movements, and linear envelopes were continuously mapped to the elbow angle by three RNN architectures: nonlinear autoregressive exogenous (NARX), Elman, and long-term short memory (LSTM). All three approaches effectively captured the complex dynamics of the multi-input to a single-output regression problem. Regarding within-subject variability, the NARX, Elman, and LSTM demonstrated superior accuracy and robustness compared to dynamic feedforward neural networks like time-delay neural networks. Notably, there was no statistically significant distinction among NARX, Elman, and LSTM estimation performances. Elman and LSTM exhibited an advantage in decoding latent information dependencies through their context layers, leading to improved estimation performance in inter-subject variability analysis, particularly with increased training data volume and variability. Furthermore, the LSTM, with its complex architecture capable of learning long-term temporal dependencies, exhibited the highest performance among the considered RNNs. Consequently, selecting the optimal RNN structure is recommended based on the complexity of the data at hand. The RNN-based decoding model holds potential applications in prosthetics, robotic assistants, and exoskeletons, enabling intention detection and real-time assessment of active rehabilitation progress. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
46. Safety management system of new energy vehicle power battery based on improved LSTM.
- Author
-
Zhao, Kun and Bai, Hao
- Subjects
METAHEURISTIC algorithms ,ELECTRIC vehicles ,FAULT diagnosis ,ENERGY management ,INDUSTRIAL efficiency - Abstract
With the development of sustainable economy, new energy materials are widely used in various industries, and many cars also adopt new energy power batteries as power sources. However, it is currently not possible to accurately diagnose faults in power batteries, which results in the safety of power batteries not being guaranteed. To address this issue, this study utilizes the Whale Optimization Algorithm to improve the Long Short-Term Memory algorithm and constructs a fault diagnosis model based on the improved algorithm. The purpose of using this model for fault diagnosis of power batteries is to strengthen the safety management of batteries. This study first conducted experiments on the improved algorithm and obtained an accuracy of 95.3%. The simulation results of the fault diagnosis model showed that the diagnosis time was only 1.2s. The analysis of the power battery showed that after using this model, the safety performance has been improved by 90.1%, while the maintenance cost has been reduced to 20.3% of the original. The above results verify that the fault diagnosis model based on the improved algorithm can accurately diagnose faults in power batteries, thereby improving the safety of power batteries. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
47. Student academic performance prediction enhancement using t-SIDSBO and Triple Voter Network.
- Author
-
Muthuselvan, S., Rajaprakash, S., Jaichandran, R., Antony, Johns, U, Amal P, and A, Ijas V
- Subjects
CONVOLUTIONAL neural networks ,RECURRENT neural networks ,DEEP learning ,ACADEMIC achievement ,PREDICTION models - Abstract
Today's educational environment considers student academic performance prediction to be extremely important in educational organizations. It is a key problem for an academic system at all educational stages. Every educational system that wants to enhance its students' studying experience and academic success must be able to forecast their academic performance. The majority of current research on predicting student performance uses traditional feature selection strategies that involve extracting characteristics and feeding them to a classifier. Scholars can now extract meaningful high-level features from unprocessed data thanks to deep learning (DL). Performance evaluation on difficult tasks is made possible by such sophisticated feature selection strategies. This work proposed a combined Triple Voter Network and t-Self Improved Distribution-based Satin Bowerbird Optimization (t-SIDSBO) predict student academic achievement. Here, the deep LSTM model, CNN model, RNN model which are based on advanced feature prediction models, is used for effective classification, and the best features are chosen using a t-SIDSBO-based feature selection strategy. In this paper, the detailed explanation of the student academic prediction and the feature selection of t-SIDSBO using DL is explained step by step procedure. Following that, the anticipated performance is assessed and improved using performance metrics like accuracy, F_ score, recall, specificity, and precision. The program is implemented using the Python platform. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
48. AN INTERPRETABLE GLAUCOMA DETECTION USING DUAL SCALE CROSS-ATTENTION VISION TRANSFORMER-BASED LONG SHORT TERM MEMORY WITH OPTICAL CUP AND DISK SEGMENTATION.
- Author
-
KRISHNAMOORTHY, V. and LOGESWARI, S.
- Subjects
- *
TRANSFORMER models , *AQUEOUS humor , *DEEP learning , *OPTIC disc , *LONG short-term memory - Abstract
Glaucoma is a kind of eye disease that tends to generate harm to the optic nerve. It is a neurodegenerative illness, which develops intraocular hypertension because of its maximized aqueous humor and blockage between the cornea and iris. It causes destruction to the optic nerve head, which transfers visual stimulus to the brain from the eyes. This results in loss of visual field and blindness. For vision, glaucoma is known to be a sneak thief due to its complexity in detecting it in the early stage. It requires continuous screening to determine the neurological disorder. Effective identification of glaucoma requires more cost and time, but it also causes human error in the detection phase based on resource availability. The problems based on the robustness of the algorithm are not solved in the earlier method especially relative to that human expert counterpart. Therefore, effective glaucoma detection with the help of deep learning is developed to recognize eye disease in the early stage. At first, the input eye images are taken from the available sites. Subsequently, the procedure for segmentation is done using the Optimized Dilated Mobile-Unet++ (ODMUnet++) to segment the optic disc and optic cup in the input images. Here, the parameters in the developed ODMUnet++ are optimized using an Improved Drawer Algorithm (IDrA). The segmented “optic disc and optic cup” images are given to the developed Dual Scale Cross-Attention Vision Transformer-based Long Short-Term Memory (DSCAViT-LSTM) for glaucoma detection. The experimental outcomes of the recommended model are evaluated with other deep learning techniques to ensure its efficacy. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
49. 基于 CF-CNN-LSTM 模型的滑坡易发性评价.
- Author
-
王守华, 王睿菘, 孙希延, 刘小明, \卢伟萍, and 子安
- Abstract
To address the problems of landslide hazard sample selection as well as long-term dependencies, gradient vanishing, and degradation in deep learning models, a CF-CNN-LSTM deep learning model is proposed. This model combines the certainty factor (CF) method, convolution neural network (CNN), and long short-term memory (LSTM) networks. Taking the Wuzhou Municipal District of Zhuang Autonomous Region, Guangxi as the study area, 13 kinds of landslide evaluation factors such as elevation, slope, slope direction, etc. were selected. The CF-CNN-LSTM model was used to evaluate the susceptibility of landslide in the study area, and compared with the CNN model, the LSTM model, the recurrent neural network model, the logistic regression model, and the prediction accuracy of the model were evaluated by six methods such as the working characteristic curve of the subjects and the overall accuracy. The results show that the AUC value of the ROC curve of the CF-CNN-LSTM model is 0.953, higher than that of other single models. At the same time, the other five assessment indexes are all better than that of the single model, which proves that the CF-CNN-LSTM model has a higher accuracy, and it can be used for landslide susceptibility evaluation in the Wuzhou City area, and it can provide scientific suggestions for landslide risk management in the region. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
50. Electricity price forecast on day-ahead market for mid- and short terms: capturing spikes in data sequences using recurrent neural network techniques.
- Author
-
Bâra, Adela and Oprea, Simona Vasilica
- Subjects
- *
RECURRENT neural networks , *ELECTRICITY pricing , *RUSSIAN invasion of Ukraine, 2022- , *PRICES , *ELECTRICITY markets - Abstract
This paper aims to forecast the electricity prices in the day-ahead market (DAM) with complex recurrent neural networks (RNNs), which are powerful in predicting the sequential prices with lags of unknown duration between significant peaks in the price curve. Recently, the electricity markets have been shaken by random events, such as the COVID-19 pandemic or the conflict in Ukraine. Therefore, long short-term memory (LSTM), Gated Recurrent Unit (GRU) and echo state networks (ESNs) are more appropriate for memorizing random events that must be remembered after some time to adequately enhance the mid- and short-run forecast. Both methods overcome the vanishing gradient problem that is common for RNN using memory cells and gates that allow the updating of the memory and tracking long-term dependencies in the input sequence. Several time series prices from neighboring East European countries and the derivation of fundamental variables are combined to predict the electricity price in Romania. The input data cover 2019–2022. The best results were obtained for 2021, whereas the best solution is provided by bi-LSTM. The prediction is proven to be reliable for the next 3–4 days. The Mean Absolute Error (MAE) almost doubled in 2022, but to further improve the results, a higher number of neurons is taken for each layer and MAE decreased. Relative to ensemble models, there was a 12.81% reduction in MAE. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.