Search

Sentiment analysis provides valuable insights into people's opinions, emotions, and attitudes, enabling businesses to make more informed decisions, improve customer satisfaction, and stay competitive in today's market. Now-a-days due to the accessibility of social networking platforms like Twitter, Instagram, Facebook, WeChat, etc a bulk of data is being generated within a very small span of time. These data inherit highly potential information hidden in them. Consequently, these data may be analyzed using powerful tools of deep learning techniques to achieve tremendous societal benefits. Keeping these views in mind, the authors have planned an extensive experimental study to undertake the task of sentiment analysis. In this present work, thirty-six different Word2Vec and GloVe Embedded deep learning models have been developed on the deep learning based architecture namely Multilayer Perceptron (MLP), Convolutional Neural Network(CNN), Long-and-Short-Term Memory Network (LSTM), Bi-directional LSTM (Bi-LSTM), Gated Recurrent Unit (GRU), and Bi-directional GRU (Bi-GRU). All the developed models are compared based on their accuracies, F1-scores, and other evaluation parameters. Several promising results have emerged from this study. It has been observed that the Word2Vec embedded Bi-directional GRU model yields the best result with an average F1-score of 0.84 in the case of a train-test ratio of 80:20. Finally this type of elaborated comparative and comprehensive study may be considered unique in its nature and supposed to be used to develop expert sentiment analyser. [ABSTRACT FROM AUTHOR]

Cyber-physical power systems (CPPS) integrate information and communication technology into conventional electric power systems to facilitate bidirectional communication of information and electric power between users and power grids. Despite its benefits, the open communication environment of CPPS is vulnerable to various security attacks. This paper proposes a federated deep learning-based architecture to detect false data injection attacks (FDIAs) in CPPS. The proposed work offers a strong, decentralized alternative with the ability to boost detection accuracy while maintaining data privacy, presenting a significant opportunity for real-world applications in the smart grid. This framework combines state-of-the-art machine learning and deep learning models, which are used in both centralized and federated learning configurations, to boost the detection of false data injection attacks in cyber-physical power systems. In particular, the research uses a multi-stage detection framework that combines several models, including classic machine learning classifiers like Random Forest and ExtraTrees Classifiers, and deep learning architectures such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU). The results demonstrate that Bidirectional GRU and LSTM models with attention layers in a federated learning setup achieve superior performance, with accuracy approaching 99.8%. This approach enhances both detection accuracy and data privacy, offering a robust solution for FDIA detection in real-world smart grid applications. [ABSTRACT FROM AUTHOR]

With the rapid development of internet and AI technologies, Agricultural Expert Systems (AESs) have become crucial for delivering technical support and decision-making in agricultural management. However, traditional natural language processing methods often struggle with specialized terminology and context, and they lack the adaptability to handle complex text classifications. The diversity and evolving nature of agricultural texts make deep semantic understanding and integration of contextual knowledge especially challenging. To tackle these challenges, this paper introduces a Bidirectional Encoder Recurrent Convolutional Neural Network (AES-BERCNN) tailored for short-text classification in agricultural expert systems. We designed an Agricultural Text Encoder (ATE) with a six-layer transformer architecture to capture both preceding and following word information. A recursive convolutional neural network based on Gated Recurrent Units (GRUs) was also developed to merge contextual information and learn complex semantic features, which are then combined with the ATE output and refined through max-pooling to form the final feature representation. The AES-BERCNN model was tested on a self-constructed agricultural dataset, achieving an accuracy of 99.63% in text classification. Its generalization ability was further verified on the Tsinghua News dataset. Compared to other models such as TextCNN, DPCNN, BiLSTM, and BERT-based models, the AES-BERCNN shows clear advantages in agricultural text classification. This work provides precise and timely technical support for intelligent agricultural expert systems. [ABSTRACT FROM AUTHOR]

Time-of-use pricing of electric energy, as an important part of the national policy of energy conservation and emission reduction, requires accurate electric energy data as support. However, due to various reasons, the electric energy data are often missing. To address this thorny problem, this paper constructs a CNN and GRU-based recovery model (RCNN-BiGRU) for electric energy data by taking the missing data as the output and the historical data of the neighboring moments as the input. Firstly, a convolutional network with a residual structure is used to capture the local dependence and periodic patterns of the input data, and then a bidirectional GRU network utilizes the extracted potential features to model the temporal relationships of the data. Aiming at the difficult selection of network structure parameters and training process parameters, an accelerated adaptive differential evolution (AADE) algorithm is proposed to optimize the electrical energy data recovery model. The algorithm designs an accelerated mutation operator and at the same time adopts an adaptive strategy to set the two key parameters. A large amount of real grid data are selected as samples to train the network, and the comparison results verify that the proposed combined model outperforms the related CNN and GRU networks. The comparison experimental results with other optimization algorithms also show that the AADE algorithm proposed in this paper has better data recovery performance on the training set and significantly better performance on the test set. [ABSTRACT FROM AUTHOR]

Deep learning algorithms have revolutionized various fields by achieving remarkable results in time series analysis. Among the different architectures, recurrent neural networks (RNNs) have played a significant role in sequential data processing. This study presents a comprehensive comparison of prominent RNN variants: long short-term memory (LSTM), Bidirectional LSTM (BiLSTM), gated recurrent unit (GRU), bidirectional GRU (BiGRU), and RNN, to analyze their respective strengths and weaknesses of national stock exchange India (NSEI). The Python application developed for this research aims to evaluate and determine the most effective algorithm among the variants. To conduct the evaluation, data from the public domain covering the period from 1/1/2004 to 30/06/2023 is collected. The dataset considers significant events such as demonetization, market crashes, the COVID-19 pandemic, downturns in the automobile sector, and rises in unemployment. Stocks from various sectors including banking, automobile, oil and gas, metal, and Pharma are selected for analysis. Finally, the results reveal that algorithm performance varies across different stocks. Specifically, in certain cases, BiLSTM outperforms, while in others, both BiGRU and LSTM are surpassed. Notably, the overall performance of simple RNN is consistently the lowest across all stocks. [ABSTRACT FROM AUTHOR]

Dialogue systems are an important research direction in artificial intelligence, with broad application prospects and market value. In order to improve system efficiency and user satisfaction, an open domain generative dialogue system integrating knowledge graphs has been developed, which facilitates the utilization of rich background knowledge during dialogue generation, thereby generating more coherent and meaningful dialogue content. At the same time, based on the sequence to sequence model, a bidirectional gated loop unit is introduced to better capture contextual information and improve the model’s understanding and generation ability. These results confirmed that the average values of the improved model in the training and validation sets were 98.66% and 87.34%, respectively, with loss values of 0.01 and 0.10. Compared to the baseline model, this improved model improved Hits@1 and Hits@3 by 0.09% and 0.25%, respectively. This improved model had the minimum perplexity of 17.62. The security and diversity of this improved system were 0.80 and 0.82, respectively, taking into account the balance of these two types of performance. Its correlation and fluency were 1.44 and 1.56, respectively. This indicates that this improved model is beneficial for improving the efficiency of generating dialogue and has certain effectiveness, better meeting users’ needs and improve user satisfaction. This system can provide users with a better conversation experience and provide technological and innovative features for artificial intelligence dialogue assistants.

Protein secondary structure prediction plays a pivotal role in deciphering protein function and structure, with implications for drug discovery, functional annotation, and molecular biology research. Deep learning techniques, particularly recurrent neural networks (RNNs), have shown promise in capturing sequential dependencies and contextual information from protein sequences. In this study, we propose a novel approach for protein secondary structure prediction by integrating Bidirectional Long Short-Term Memory (BiLSTM) and Bidirectional Gated Recurrent Unit (BiGRU) architectures. Leveraging the complementary strengths of BiLSTM and BiGRU networks, our integrated model aims to enhance prediction accuracy and robustness by effectively capturing both short and long-range dependencies within protein sequences. We evaluate the performance of the proposed model on benchmark datasets and compare it with state-of-the-art methods in the field. The model is trained on CB6133 filtered dataset and tested on CB513, CASP13 and CASP14 dataset. Our results demonstrate 87.92% and 78.6% Q3 and Q8 accuracy on CB513 dataset.

Abstract Cardiac disease has become a severe threat to public health according to the government report. In China, there are 0.29 billion cardiac patients and early diagnosis will greatly reduce mortality and improve life quality. Electrocardiogram (ECG) signal is a priority tool in the diagnosis of heart diseases because it is non‐invasive and easily available with a simple diagnostic tool of low cost. The paper proposes an automatic classification model by combing convolutional neural network (CNN) and recurrent neural network (RNN) to distinguish different types of cardiac arrhythmias. Morphology features of the raw ECG signals are extracted by CNN blocks and fed into a bidirectional gated recurrent unit (GRU) network. Attention mechanism is used to highlight specific features of the input sequence and contribute to the performance improvement of classification. The model is evaluated with two datasets considering the class imbalance problem constructed with records from MIT‐BIH arrhythmia database and China Physiological Signal Challenge 2018 database. Experimental results show that this model achieves good performance with an average F1 score of 0.9110 on public dataset and 0.9082 on subject‐specific dataset, which may have potential practical applications.

Abstract The adverse effect of air pollution has always been a problem for human health. The presence of a high level of air pollutants can cause severe illnesses such as emphysema, chronic obstructive pulmonary disease (COPD), or asthma. Air quality prediction helps us to undertake practical action plans for controlling air pollution. The Air Quality Index (AQI) reflects the degree of concentration of pollutants in a locality. The average AQI was calculated for the various cities in China to understand the annual trends. Furthermore, the air quality index has been predicted for ten major cities across China using five different deep learning techniques, namely, Recurrent Neural Network (RNN), Bidirectional Gated Recurrent unit (Bi-GRU), Bidirectional Long Short-Term Memory (BiLSTM), Convolutional Neural Network BiLSTM (CNN-BiLSTM), and Convolutional BiLSTM (Conv1D-BiLSTM). The performance of these models has been compared with a machine learning model, eXtreme Gradient Boosting (XGBoost) to discover the most efficient deep learning model. The results suggest that the machine learning model, XGBoost, outperforms the deep learning models. While Conv1D-BiLSTM and CNN-BiLSTM perform well among the deep learning models in the estimation of the air quality index (AQI), RNN and Bi-GRU are the least performing ones. Thus, both XGBoost and neural network models are capable of capturing the non-linearity present in the dataset with reliable accuracy.

Finding a reliable approach to detect bearing faults is crucial, as the most common rotating machine defects occur in its bearings. A convolutional neural network can automatically extract the local features of the mechanical vibration signal and classify the patterns. Nevertheless, these types of networks suffer from the extraction of the global feature of the input signal as they utilize only one scale on their input. This paper presents a method to overcome the above weakness by employing a combination of three parallel convolutional neural networks with different filter lengths. In addition, a bidirectional gated recurrent unit is utilized to extract global features. The CWRU-bearing dataset is used to prove the performance of the proposed method. The results show the high accuracy of the proposed method even in the presence of noise.

For automatic and intelligent check of operation ticket for power grid scheduling， a scheduling operation ticket checking and analysis method based on bidirectional GRU （gated recurrent unit） neural networks and multiple verification rules is proposed. The semantic analysis technology based on bidirectional GRU neural network is combined with the intelligent checking method for data pre-processing of dispatching instruction and maintenance application form. The state of safety protection measures is compared with the final state of operation ticket equipment. In the intelligent rule checking link， format checking， logic checking and safety checking are performed in turn， and finally the ticket errors are output. The proposed method is tested using the historical dispatching operation tickets in the intelligent outage control system in Zhejiang power grid. The results show that the method can improve the checking safety and efficiency and realize the intelligent operation ticket checking of power grid dispatching.

Copyright of Journal of Computer Science & Technology (JCS&T) is the property of Journal of Computer Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The adverse effect of air pollution has always been a problem for human health. The presence of a high level of air pollutants can cause severe illnesses such as emphysema, chronic obstructive pulmonary disease (COPD), or asthma. Air quality prediction helps us to undertake practical action plans for controlling air pollution. The Air Quality Index (AQI) reflects the degree of concentration of pollutants in a locality. The average AQI was calculated for the various cities in China to understand the annual trends. Furthermore, the air quality index has been predicted for ten major cities across China using five different deep learning techniques, namely, Recurrent Neural Network (RNN), Bidirectional Gated Recurrent unit (Bi-GRU), Bidirectional Long Short-Term Memory (BiLSTM), Convolutional Neural Network BiLSTM (CNN-BiLSTM), and Convolutional BiLSTM (Conv1D-BiLSTM). The performance of these models has been compared with a machine learning model, eXtreme Gradient Boosting (XGBoost) to discover the most efficient deep learning model. The results suggest that the machine learning model, XGBoost, outperforms the deep learning models. While Conv1D-BiLSTM and CNN-BiLSTM perform well among the deep learning models in the estimation of the air quality index (AQI), RNN and Bi-GRU are the least performing ones. Thus, both XGBoost and neural network models are capable of capturing the non-linearity present in the dataset with reliable accuracy. [ABSTRACT FROM AUTHOR]

The mood is a psychological state of feeling that is related to internal emotions and affect, which is how emotions are expressed outwardly. According to studies, music affects our moods, and we are also inclined to choose a theme based on our current moods. Audio-based techniques can achieve promising results, but lyrics also give relevant information about the moods of a song which may not be present in the audio part. So a multi-modal with both textual features and acoustic features can provide enhanced accuracy. Sequential networks such as long short-term memory networks (LSTM) and gated recurrent unit networks (GRU) are widely used in the most state-of-the-art natural language processing (NLP) models. A transformer model uses self-attention to compute representations of its inputs and outputs, unlike recurrent unit networks (RNNs) that use sequences and transformers that can parallelize over input positions during training. In this work, we proposed a multi-modal music mood classification system based on transformers and compared the system's performance using a bi-directional GRU (Bi-GRU)-based system with and without attention. The performance is also analyzed for other state-of-the-art approaches. The proposed transformer-based model acquired higher accuracy than the Bi-GRU-based multi-modal system with single-layer attention by providing a maximum accuracy of 77.94\%.

Handwriting recognition is a challenge that interests many researchers around the world. As an exception, handwritten Arabic script has many objectives that remain to be overcome, given its complex form, their number of forms which exceeds 100 and its cursive nature. Over the past few years, good results have been obtained, but with a high cost of memory and execution time. In this paper we propose to improve the capacity of bidirectional gated recurrent unit (BGRU) to recognize Arabic text. The advantages of using BGRUs is the execution time compared to other methods that can have a high success rate but expensive in terms of time andmemory. To test the recognition capacity of BGRU, the proposed architecture is composed by 6 convolutional neural network (CNN) blocks for feature extraction and 1 BGRU + 2 dense layers for learning and test. The experiment is carried out on the entire database of institut für nachrichtentechnik/ecole nationale d'ingénieurs de Tunis (IFN/ENIT) without any preprocessing or data selection. The obtained results show the ability of BGRUs to recognize handwritten Arabic script. [ABSTRACT FROM AUTHOR]

Copyright of Zhejiang Electric Power is the property of Zhejiang Electric Power Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Deep learning models in recent times have shown promising results for solar energy forecasting. Solar energy depends heavily on local weather conditions, and as a result, typically hundreds of models are built, which need site and season-specific training. The model maintenance and management also become a tedious job with such a large number of models. Here, we are motivated to use transfer learning to accommodate local variations in the solar pattern over the available global pattern. It may also be noted that apparently transfer learning has been rarely/never used for solar forecasting. In this paper, we have proposed a bidirectional gated recurrent unit (BGRU) based model, which employs transfer learning for short-term solar energy forecasting. The said model yields better forecasting accuracy compared to site-specific models with a lower variance. It also takes 39.6% less parameters and 76.1% reduced time for training. The current literature suggests that selection of base scenario for transfer learning is an open problem and in this paper, we have also proposed an intuitive strategy for the same. The effectiveness of the same is established through empirical study. [ABSTRACT FROM AUTHOR]

Complex weather conditions—in particular clouds—leads to uncertainty in photovoltaic (PV) systems, which makes solar energy prediction very difficult. Currently, in the renewable energy domain, deep-learning-based sequence models have reported better results compared to state-of-the-art machine-learning models. There are quite a few choices of deep-learning architectures, among which Bidirectional Gated Recurrent Unit (BGRU) has apparently not been used earlier in the solar energy domain. In this paper, BGRU was used with a new augmented and bidirectional feature representation. The used BGRU network is more generalized as it can handle unequal lengths of forward and backward context. The proposed model produced 59.21%, 37.47%, and 76.80% better prediction accuracy compared to traditional sequence-based, bidirectional models, and some of the established states-of-the-art models. The testbed considered for evaluation of the model is far more comprehensive and reliable considering the variability in the climatic zones and seasons, as compared to some of the recent studies in India.

Network traffic matrix prediction is a methodology of predicting network traffic behavior ahead of time in order to improve network management and planning. Different neural network models ranging from simple recurrent neural network (RNN) to long short-term memory neural network (LSTM) and gated recurrent unit (GRU) are being used to predict traffic matrix. In this paper, for the first time the bidirectional LSTM (Bi-LSTM) and the bidirectional GRU (Bi-GRU) are applied to predict the network traffic matrix due to their high effectiveness and efficiency. The proposed models were designed as hybrid models that support multiple neural network models in a chained manner to support higher feature learning and subsequently higher accuracies in traffic matrix prediction. The hybrid models combined convolutional neural network (CNN) with either Bi-LSTM or Bi-GRU along with the unidirectional versions. With this approach, it gives the ability to eliminate unneeded information in order to obtain good data prediction. The comparisons of the proposed methods were applied on real traffic data from the GÉANT network. The results showed that the proposed models have a considerable improvement in prediction accuracy when compared to other existing models found in literature.

In an increasingly open electricity market environment, short-term load forecasting (STLF) can ensure the power grid to operate safely and stably, reduce resource waste, power dispatching, and provide technical support for demand-side response. Recently, with the rapid development of demand side response, accurate load forecasting can better provide demand side incentive for regional load of prosumer groups. Traditional machine learning prediction and time series prediction based on statistics failed to consider the non-linear relationship between various input features, resulting in the inability to accurately predict load changes. Recently, with the rapid development of deep learning, extensive research has been carried out in the field of load forecasting. On this basis, a feature selection algorithm based on random forest is first used in this paper to provide a basis for the selection of the input features of the load forecasting model. After the input features are selected, a hybrid neural network STLF algorithm based on multi-model fusion is proposed, of which the main structure of the hybrid neural network is composed of convolutional neural network and bidirectional gated recurrent unit (CNN-BiGRU). The input data is obtained by using long sliding time windows of different steps, then multiple CNN-BiGRU models are trained respectively. The forecasting results of multiple models are averaged to get the final forecasting load value. The load datasets come from a region in New Zealand and a region in Zhejiang, China, are used as load forecast examples. Finally, a variety of load forecasting algorithms are introduced for comparison. The experimental results show that our method has a higher accuracy than comparison models.

Traditional aerial target tactical intention recognition is based on a single moment of reasoning, while actual battlefield target tactical intention is realized by a series of actions, so the target state reflects dynamic and temporal variation. To solve this problem, bidirectional propagation and attention mechanisms are introduced based on a gated recurrent unit (GRU) network, and bidirectional gated recurrent units with attention mechanism (BiGRU-Attention) air target tactical intention recognition model is proposed. We use a hierarchical approach to construct the air combat intention characteristic set, encode it into temporal characteristics, encapsulate the decision-maker’s experience into labels, learn the deep-level information in the air combat intention characteristic vector through a BiGRU neural network, and use the attention mechanism to adaptively assign network weights, and then place air combat characteristic information with different weights in a softmax function layer for intention recognition. Comparison with a traditional air tactical target intention recognition model and analysis of ablation experiments show that the proposed model effectively improves the tactical intention recognition of air targets.

Sentiment analysis of e-commerce reviews is the hot topic in the e-commerce product quality management, from which manufacturers are able to learn the public sentiment about products being sold on e-commerce websites. Meanwhile, customers can know other people's attitudes about the same products. This paper proposes the deep learning model of Bert-BiGRU-Softmax with hybrid masking, review extraction and attention mechanism, which applies sentiment Bert model as the input layer to extract multi-dimensional product feature from e-commerce reviews, Bidirectional GRU model as the hidden layer to obtain semantic codes and calculate sentiment weights of reviews, and Softmax with attention mechanism as the output layer to classify the positive or negative nuance. A series of experiments are conducted on the large-scale dataset involving over 500 thousand product reviews. The results show that the proposed model outperforms the other deep learning models, including RNN, BiGRU, and Bert-BiLSTM, which can reach over 95.5% of accuracy and retain a lower loss for the e-commerce reviews.

Entity typing (ET) is the process of identifying the semantic types of every entity within a corpus. ET involves labelling each entity mention with one or more class labels. As a multi-class, multi-label task, it is considerably more challenging than named entity recognition. This means existing entity typing models require pre-identified mentions and cannot operate directly on plain text. Pipeline-based approaches are therefore used to join a mention extraction model and an entity typing model to process raw text. Another key limiting factor is that these mention-level ET models are trained on fixed context windows, which makes the entity typing results sensitive to window size selection. In light of these drawbacks, we propose an end-to-end entity typing model (E2EET) using a Bi-GRU to remove the dependency on window size. To demonstrate the effectiveness of our E2EET model, we created a stronger baseline mention-level model by incorporating the latest contextualised transformer-based embeddings (BERT). Extensive ablative studies demonstrate the competitiveness and simplicity of our end-to-end model for entity typing. [ABSTRACT FROM AUTHOR]

Complex weather conditions--in particular clouds--leads to uncertainty in photovoltaic (PV) systems, which makes solar energy prediction very difficult. Currently, in the renewable energy domain, deep-learning-based sequence models have reported better results compared to state-of-theart machine-learning models. There are quite a few choices of deep-learning architectures, among which Bidirectional Gated Recurrent Unit (BGRU) has apparently not been used earlier in the solar energy domain. In this paper, BGRU was used with a new augmented and bidirectional feature representation. The used BGRU network is more generalized as it can handle unequal lengths of forward and backward context. The proposed model produced 59.21%, 37.47%, and 76.80% better prediction accuracy compared to traditional sequence-based, bidirectional models, and some of the established states-of-the-art models. The testbed considered for evaluation of the model is far more comprehensive and reliable considering the variability in the climatic zones and seasons, as compared to some of the recent studies in India. [ABSTRACT FROM AUTHOR]

E-commerce websites produce a large number of online reviews, posts, and comments about a product or service. These reviews are used to assist consumers in buying or recommending a product. However, consumers are expressing their views on a specific aspect category of a product. In particular, aspect category detection is one of the subtasks of aspect-based sentiment analysis, and it classifies a given text into a set of predefined aspects. Naturally, a class imbalance problem occurs in real-world applications. The class imbalance is studied over the last two decades using machine learning algorithms. However, there is very little empirical research in deep learning with the class imbalance problem. In this paper, we propose bidirectional LSTM and GRU networks to deal with imbalance aspect categories. The proposed method applies a data-level technique to reduce class imbalance. Specifically, we employ the stratified sampling technique to deal with imbalanced classes. Moreover, we create word vectors with the corpus-specific word embeddings and pre-trained word embeddings. This word representations fed into the proposed method and their merge modes such as addition, multiplication, average, and concatenation. The performance of this method is evaluated with a confusion matrix, precision, recall, F1-score with micro-average, macro-average, and weighted average. The experimental result analysis suggests that the proposed method outperforms with pre-trained word embeddings. [ABSTRACT FROM AUTHOR]

Traditional humor computing tasks rely on artificially constructed features, which are prone to loss of features, and mainly focus on humorous judgment. The Siamese bidirectional GRU attention model based on deep learning compared pairs of humorous texts to determine which sentence was more humorous. Firstly, this paper used the text processor to train the word embedding. Secondly, it used the bidirectional GRU model to obtain the annotation of each word. Finally, it performed the humorous comparison task at the fully connected layer. This paper conducted experiments on the Semeval-2017 Task6-#HashtagWars data set, and used accuracy as an evaluation indicator. The experimental results show that the model and other related models have a significant improvement in the comparison of humorous text. [ABSTRACT FROM AUTHOR]

In this paper, we propose a hierarchical bidirectional Gated Recurrent Unit (GRU) network with attention for human emotion classification from continues electroencephalogram (EEG) signals. The structure of the model mirrors the hierarchical structure of EEG signals, and the attention mechanism is used at two levels of EEG samples and epochs. By paying different levels of attention to content with different importance, the model can learn more significant feature representation of EEG sequence which highlights the contribution of important samples and epochs to its emotional categories. We conduct the cross-subject emotion classification experiments on DEAP data set to evaluate the model performance. The experimental results show that in valence and arousal dimensions, our model on 1-s segmented EEG sequences outperforms the best deep baseline LSTM model by 4.2% and 4.6%, and outperforms the best shallow baseline model by 11.7% and 12% respectively. Moreover, with increase of the epoch's length of EEG sequences, our model shows more robust classification performance than baseline models, which demonstrates that the proposed model can effectively reduce the impact of long-term non-stationarity of EEG sequences and improve the accuracy and robustness of EEG-based emotion classification.

Network traffic matrix prediction is a methodology of predicting network traffic behavior ahead of time in order to improve network management and planning. Different neural network models ranging from simple recurrent neural network (RNN) to long short-term memory neural network (LSTM) and gated recurrent unit (GRU) are being used to predict traffic matrix. In this paper, for the first time the bidirectional LSTM (Bi-LSTM) and the bidirectional GRU (Bi-GRU) are applied to predict the network traffic matrix due to their high effectiveness and efficiency. The proposed models were designed as hybrid models that support multiple neural network models in a chained manner to support higher feature learning and subsequently higher accuracies in traffic matrix prediction. The hybrid models combined convolutional neural network (CNN) with either Bi-LSTM or Bi-GRU along with the unidirectional versions. With this approach, it gives the ability to eliminate unneeded information in order to obtain good data prediction. The comparisons of the proposed methods were applied on real traffic data from the GÉANT network. The results showed that the proposed models have a considerable improvement in prediction accuracy when compared to other existing models found in literature. [ABSTRACT FROM AUTHOR]

• A framework based on Bidirectional Gated Neural Networks is proposed for fault diagnosis with uncertainty in dynamic environments. • A Sample Sensitive Bidirectional Gated Neural Networks model is developed to tackle imbalanced fault diagnosis challenges. • Cost sensitive active learning is used to explore unlabeled data. • Effective methods are developed to address both binary Fault Diagnosis and multi-class Fault Diagnosis. Most existing fault diagnosis methods may fail in the following three scenarios: (1) serial correlations exist in the process data; (2) fault data are much less than normal data; and (3) it is impractical to obtain enough labeled data. In this paper, a novel form of the bidirectional gated recurrent unit (BGRU) is developed to underpin effective and efficient fault diagnosis using cost sensitive active learning. Specifically, BGRU is devised to consider the dynamic behavior of a complex process. In the training phase of BGRU, the idea of weighting each training example is proposed to reduce the effect of class imbalance. Besides, in order to explore the unlabeled data, cost sensitive active learning is utilized to select the candidate instances. The effectiveness of the proposed method is evaluated on the Tennessee Eastman (TE) dataset and a real plasma etching process dataset. The experiment results show that the proposed cost senstive active learning bidirectional gated recurrent unit (CSALBGRU) method achieves better performance in both binary fault diagnosis and multi-class fault diagnosis. [ABSTRACT FROM AUTHOR]

The user’s feedback on healthcare services is usually based on ratings from post-service questionnaires. However, in order to get a clear view of the user's perspective, online text reviews need to be analyzed. We combined targeted and aspect-based sentiment analysis by multi-level attention to get a specific user sentiment on a target of an aspect. The multi-level attention consists of Target-level and Sentence-level attention. Our proposed framework is based on Bidirectional Gated Recurrent Unit. Bi-GRU is commonly known to have comparable results compared to LSTM while having lesser computational complexity. We also utilized Bidirectional LSTM based Character-Enhanced Token-Embedding to handle out of vocabulary words and misspelling to avoid error in detecting sentiment. We created a dataset of online healthcare reviews from 2018-2020, targeting the name of the hospital or department, with ten aspects: cleanliness, cost, doctor, food, nurse, parking, receptionist and billing, safety, test and examination, and waiting time. To improve the results of our proposed method, we calculated polarity weight to handle imbalanced aspects in the dataset. We classified these reviews into three polarities, which are positive, negative, and neutral. Based on our experiments, we achieved the best F1-Score of 88%. [ABSTRACT FROM AUTHOR]

RNA-Protein binding plays important roles in the field of gene expression. With the development of high throughput sequencing, several conventional methods and deep learning-based methods have been proposed to predict the binding preference of RNA-protein binding. These methods can hardly meet the need of consideration of the dependencies between subsequence and the various motif lengths of different translation factors (TFs). To overcome such limitations, we propose a predictive model that utilizes a combination of multi-scale convolutional layers and bidirectional gated recurrent unit (GRU) layer. Multi-scale convolution layer has the ability to capture the motif features of different lengths, and bidirectional GRU layer is able to capture the dependencies among subsequence. Experimental results show that the proposed method performs better than four state-of-the-art methods in this field. In addition, we investigate the effect of model structure on model performance by performing our proposed method with a different convolution layer and a different number of kernel size. We also demonstrate the effectiveness of bidirectional GRU in improving model performance through comparative experiments. [ABSTRACT FROM AUTHOR]

Accurate and robust estimation of Remaining Useful life (RUL) is of paramount importance for development of advanced smart and predictive maintenance strategies. To this aim, the paper proposes a new hybrid framework, referred to as the NPBGRU, developed by integration of three fully noisy deep learning architectures. Noisy CNN (NCNN) and Noisy Bi-directional GRU (NBGRU) paths are designed in parallel and their concatenated output is fed into the Noisy fusion center (NFC). Adopting the proposed noisy layers enhances the robustness and generalization behavior of the proposed model. The proposed NPBGRU framework is validated using NASA's C-MAPSS dataset, illustrating state-of-the-art results. [ABSTRACT FROM AUTHOR]

Neural networks lately have achieved a great success on sentiment classification due to their ability of feature extraction. However, it remains as an enormous challenge to model long texts in document-level sentiment classification as well as to explore semantics between sentences and dependencies between features. Moreover, most existing methods can hardly differentiate the importance of different contents when constructing a document representation. To tackle these problems, we propose a novel neural network model: AttDR-2DCNN, which mainly consists of two parts. The first part utilizes a two-layer compositional bidirectional Gated Recurrent Unit (GRU) to obtain the compositional semantics of the document, where the first layer learns the feature vector of the sentence, and the second layer learns the document matrix representation consisting of two dimensions of the time-step dimension and feature dimension, from the sentence vectors. The second part applies a two-dimensional convolution operation and two-dimensional max pooling to capture more dependencies between sentences features. We further utilize different types of attention mechanism in these two parts to distinguish the importance of words, sentences and features in the document. Experiments are conducted on four publicly available document-level review datasets and the result shows that the proposed model outperforms some existing models. [ABSTRACT FROM AUTHOR]

Automatic music recommendation systems based on human emotions are becoming popular nowadays. Since audio and lyrics can provide a rich set of information regarding a song, a fusion model including both modalities can enhance classification accuracy and is attempted in this paper. The main objective of the paper is to address music mood classification using various attention mechanisms, namely, self-attention (SA), channel attention (CA), and hierarchical attention network (HAN), on a multi-modal music mood classification system. Experimental results show that multi-modal architectures with attention have achieved higher accuracy than multi-modal architectures without attention and uni-modal architectures. Motivated by the performance of attention mechanisms, a new network architecture, HAN-CA-SA based multi-modal classification system, is proposed, which reported an accuracy of 82.35%. ROC and Kappa are also computed to see the efficacy of the proposed model. The proposed model is also evaluated using the K-fold cross-validation technique. The performance of the proposed model is compared with that of XLNet and CNN-BERT systems. In addition, McNemar's statistical hypothesis test is conducted to reaffirm the importance of the proposed approach. [ABSTRACT FROM AUTHOR]

Humans have traditionally found it easy to determine emotions from facial expressions, but doing so by utilizing a computer algorithm is far more challenging. It is now feasible to discern emotions from photographs because of recent advances in computer vision as well as artificial intelligence. The following issues afflict existing face expression detection systems throughout the application process: As the extant shallow feature extraction framework has lost a great deal of effective feature data and also has limited detection performance, the paper proposes a novel technique termed "Improved Deep CNN-based Two Stream Super Resolution and Hybrid Deep Model-based Facial Emotion Detection", which consists of three working phases: super-resolution, facial emotion recognition, as well as classification. Improved Deep CNN has been used in the super-resolution phase for two streams, structure and texture stream super-resolution, that delivers high pixel density to the pictures with minimal cross-entropy loss. The facial emotion identification phase involves face recognition utilizing the Viola–Jones face detection algorithm and feature extraction using three traditional methods: Texton, Bag of Words (BOW), and GLCM, as well as improved LGXP features. In addition, classification was performed in our work utilizing RNN as well as Bi-GRU neural networks. A voting mechanism known as Score level fusion was used to get precise categorization results. Our unique Deep CNN-based facial emotion recognition system provides 95% accuracy, which is superior to other traditional approaches, according to the results. Positive and negative performance metrics for various databases have also been assessed and compared with our proposed methodology, proving that our novel strategy surpasses conventional approaches. [ABSTRACT FROM AUTHOR]

• The electrical load series is analyzed in a two dimensional matrix form. • Univariate load and temperature data are expanded to multidimensional features. • 2-D convolutional layers are used for hidden feature extraction. • A network containing LSTM and GRU units is proposed for load forecasting. The consumed electrical load is affected by many external factors such as weather, season of the year, weekday or weekend and holiday. In this paper, it is tried to provide a high accurate forecasting model for hourly load consumption with considering these external variables. At first, the electrical load and temperature time series are rearranged into separate two-dimensional matrices. Convolutional neural networks (CNNs) are utilized to extract the load and temperature features. The autocorrelation coefficients of the load and temperature sequences are used to determine the kernel size of the convolutional layers. At this stage, the convolutional layers specifically convert the univariate data to multidimensional features by applying two-dimensional convolutional kernels, which potentially increase the forecasting capability of recurrent neural networks. On the other hand, long short term memory (LSTM) and gated recurrent unit (GRU) are able to hold short-term and long-term memories. Therefore, in the next stage, the multidimensional features extracted by 2-D CNNs are fed as input to the bidirectional propagating GRU and LSTM units to perform hourly electrical load forecasting. The results of experiments on two datasets show the superiority of the proposed method compared to some recent works in the field of short-term load forecasting. [ABSTRACT FROM AUTHOR]