Search

Deep random forest (DRF), which combines deep learning and random forest, exhibits comparable accuracy, interpretability, low memory and computational overhead to deep neural networks (DNNs) in edge intelligence tasks. However, efficient DRF accelerator is lagging behind its DNN counterparts. The key to DRF acceleration lies in realizing the branch-split operation at decision nodes. In this work, we propose implementing DRF through associative searches realized with ferroelectric analog content addressable memory (ACAM). Utilizing only two ferroelectric field effect transistors (FeFETs), the ultra-compact ACAM cell performs energy-efficient branch-split operations by storing decision boundaries as analog polarization states in FeFETs. The DRF accelerator architecture and its model mapping to ACAM arrays are presented. The functionality, characteristics, and scalability of the FeFET ACAM DRF and its robustness against FeFET device non-idealities are validated in experiments and simulations. Evaluations show that the FeFET ACAM DRF accelerator achieves ~106x/10x and ~106x/2.5x improvements in energy and latency, respectively, compared to other DRF hardware implementations on state-of-the-art CPU/ReRAM. [ABSTRACT FROM AUTHOR]

The Yellow River Delta wetlands in China belong to the coastal wetland ecosystem, which is one of the youngest and most characteristic wetlands in the world. The Yellow River Delta wetlands are constantly changed by inland sediment and the influence of waves and storm surges, so the accurate classification of the coastal wetlands in the Yellow River Delta is of great significance for the rational utilization, development and protection of wetland resources. In this study, the Yellow River Delta sentinel-2 multispectral data were processed by super-resolution synthesis, and the feature bands were optimized. The optimal feature-band combination scheme was screened using the OIF algorithm. A deep learning model attention mechanism ResNet based on feature optimization with attention mechanism integration into the ResNet network is proposed. Compared with the classical machine learning model, the AM_ResNet model can effectively improve the classification accuracy of the wetlands in the Yellow River Delta. The overall accuracy was 94.61% with a Kappa of 0.93, and they were improved by about 6.99% and 0.1, respectively, compared with the best-performing Random Forest Classification in machine learning. The results show that the method can effectively improve the classification accuracy of the wetlands in the Yellow River Delta. [ABSTRACT FROM AUTHOR]

Skin cancer has a significant impact on the lives of many individuals annually and is recognized as the most prevalent type of cancer. In the United States, an estimated annual incidence of approximately 3.5 million people receiving a diagnosis of skin cancer underscores its widespread prevalence. Furthermore, the prognosis for individuals afflicted with advancing stages of skin cancer experiences a substantial decline in survival rates. This paper is dedicated to aiding healthcare experts in distinguishing between benign and malignant skin cancer cases by employing a range of machine learning and deep learning techniques and different feature extractors and feature selectors to enhance the evaluation metrics. In this paper, different transfer learning models are employed as feature extractors, and to enhance the evaluation metrics, a feature selection layer is designed, which includes diverse techniques such as Univariate, Mutual Information, ANOVA, PCA, XGB, Lasso, Random Forest, and Variance. Among transfer models, DenseNet-201 was selected as the primary feature extractor to identify features from data. Subsequently, the Lasso method was applied for feature selection, utilizing diverse machine learning approaches such as MLP, XGB, RF, and NB. To optimize accuracy and precision, ensemble methods were employed to identify and enhance the best-performing models. The study provides accuracy and sensitivity rates of 87.72% and 92.15%, respectively. [ABSTRACT FROM AUTHOR]

Monitoring snow avalanche activity is essential for operational avalanche forecasting and the successful implementation of mitigation measures to ensure safety in mountain regions. To facilitate and automate the monitoring process, avalanche detection systems equipped with seismic sensors can provide a cost-effective solution. Still, automatically differentiating avalanche signals from other sources in seismic data remains challenging, mainly due to the complexity of seismic signals generated by avalanches, the complex signal transmission through the ground, the relatively rare occurrence of avalanches, and the presence of multiple sources in the continuous seismic data. One approach to automate avalanche detection is by applying machine learning methods. So far, research in this area has mainly focused on extracting standard domain-specific signal attributes in the time and frequency domains as input features for statistical models. In this study, we propose a novel application of deep learning autoencoder models for the automatic and unsupervised extraction of features from seismic recordings. These new features are then fed into classifiers for discriminating snow avalanches. To this end, we trained three Random forest classifiers based on different feature extraction approaches. The first set of 32 features was automatically extracted from the time-series signals by an autoencoder consisting of convolutional layers and a recurrent long short-term memory unit. The second autoencoder applies a series of fully connected layers to extract 16 features from the spectrum of the signals. As a benchmark, a third random forest was trained with typical waveform, spectral and spectrogram attributes used to discriminate seismic events. We extracted all these features from 10-second windows of the seismograms recorded with an array of five seismometers installed in an avalanche test site located above Davos, Switzerland. The database used to train and test the models contained 84 avalanches and 828 noise (unrelated to avalanches) events recorded during the winter seasons of 2020–2021 and 2021–2022. Finally, we assessed the performance of each classifier, compared the results, and proposed different aggregation methods to improve the predictive performance of the developed seismic detection algorithms. The classifiers achieved an avalanche f1-score of 0.61 (seismic attributes), 0.49 (temporal autoencoder) and 0.60 (spectral autoencoder) and avalanche recall of 0.68, 0.71 and 0.71, respectively. Overall, the macro f1-score ranged from 0.70 (temporal autoencoder) to 0.78 (seismic attributes). After applying a post-processing step to event-based predictions, the avalanche recall of the three models significantly increased, reaching values between 0.82 and 0.91. The developed approach could be potentially used as an operational, near-real-time avalanche detection system. Yet, the relatively high number of false alarms still needs further implementation of the current automated seismic classification algorithms to be used as unique methods to detect avalanches effectively. [ABSTRACT FROM AUTHOR]

Due to the recent advances in the Internet and communication technologies, network systems and data have evolved rapidly. The emergence of new attacks jeopardizes network security and make it really challenging to detect intrusions. Multiple network attacks by an intruder are unavoidable. Our research targets the critical issue of class imbalance in intrusion detection, a reflection of the real-world scenario where legitimate network activities significantly out number malicious ones. This imbalance can adversely affect the learning process of predictive models, often resulting in high false-negative rates, a major concern in Intrusion Detection Systems (IDS). By focusing on datasets with this imbalance, we aim to develop and refine advanced algorithms and techniques, such as anomaly detection, cost-sensitive learning, and oversampling methods, to effectively handle such disparities. The primary goal is to create models that are highly sensitive to intrusions while minimizing false alarms, an essential aspect of effective IDS. This approach is not only practical for real-world applications but also enhances the theoretical understanding of managing class imbalance in machine learning. Our research, by addressing these significant challenges, is positioned to make substantial contributions to cybersecurity, providing valuable insights and applicable solutions in the fight against digital threats and ensuring robustness and relevance in IDS development. An intrusion detection system (IDS) checks network traffic for security, availability, and being non-shared. Despite the efforts of many researchers, contemporary IDSs still need to further improve detection accuracy, reduce false alarms, and detect new intrusions. The mean convolutional layer (MCL), feature-weighted attention (FWA) learning, a bidirectional long short-term memory (BILSTM) network, and the random forest algorithm are all parts of our unique hybrid model called MCL-FWA-BILSTM. The CNN-MCL layer for feature extraction receives data after preprocessing. After convolution, pooling, and flattening phases, feature vectors are obtained. The BI-LSTM and self-attention feature weights are used in the suggested method to mitigate the effects of class imbalance. The attention layer and the BI-LSTM features are concatenated to create mapped features before feeding them to the random forest algorithm for classification. Our methodology and model performance were validated using NSL-KDD and UNSW-NB-15, two widely available IDS datasets. The suggested model's accuracies on binary and multi-class classification tasks using the NSL-KDD dataset are 99.67% and 99.88%, respectively. The model's binary and multi-class classification accuracies on the UNSW-NB15 dataset are 99.56% and 99.45%, respectively. Further, we compared the suggested approach with other previous machine learning and deep learning models and found it to outperform them in detection rate, FPR, and F-score. For both binary and multiclass classifications, the proposed method reduces false positives while increasing the number of true positives. The model proficiently identifies diverse network intrusions on computer networks and accomplishes its intended purpose. The suggested model will be helpful in a variety of network security research fields and applications. [ABSTRACT FROM AUTHOR]

Automated disease diagnosis and prediction, powered by AI, play a crucial role in enabling medical professionals to deliver effective care to patients. While such predictive tools have been extensively explored in resource-rich languages like English, this manuscript focuses on predicting disease categories automatically from symptoms documented in the Afaan Oromo language, employing various classification algorithms. This study encompasses machine learning techniques such as support vector machines, random forests, logistic regression, and Naïve Bayes, as well as deep learning approaches including LSTM, GRU, and Bi-LSTM. Due to the unavailability of a standard corpus, we prepared three data sets with different numbers of patient symptoms arranged into 10 categories. The two feature representations, TF-IDF and word embedding, were employed. The performance of the proposed methodology has been evaluated using accuracy, recall, precision, and F1 score. The experimental results show that, among machine learning models, the SVM model using TF-IDF had the highest accuracy and F1 score of 94.7%, while the LSTM model using word2vec embedding showed an accuracy rate of 95.7% and F1 score of 96.0% from deep learning models. To enhance the optimal performance of each model, several hyper-parameter tuning settings were used. This study shows that the LSTM model verifies to be the best of all the other models over the entire dataset. [ABSTRACT FROM AUTHOR]

Recently, explainability in machine and deep learning has become an important area in the field of research as well as interest, both due to the increasing use of artificial intelligence (AI) methods and understanding of the decisions made by models. The explainability of artificial intelligence (XAI) is due to the increasing consciousness in, among other things, data mining, error elimination, and learning performance by various AI algorithms. Moreover, XAI will allow the decisions made by models in problems to be more transparent as well as effective. In this study, models from the 'glass box' group of Decision Tree, among others, and the 'black box' group of Random Forest, among others, were proposed to understand the identification of selected types of currant powders. The learning process of these models was carried out to determine accuracy indicators such as accuracy, precision, recall, and F1-score. It was visualized using Local Interpretable Model Agnostic Explanations (LIMEs) to predict the effectiveness of identifying specific types of blackcurrant powders based on texture descriptors such as entropy, contrast, correlation, dissimilarity, and homogeneity. Bagging (Bagging_100), Decision Tree (DT0), and Random Forest (RF7_gini) proved to be the most effective models in the framework of currant powder interpretability. The measures of classifier performance in terms of accuracy, precision, recall, and F1-score for Bagging_100, respectively, reached values of approximately 0.979. In comparison, DT0 reached values of 0.968, 0.972, 0.968, and 0.969, and RF7_gini reached values of 0.963, 0.964, 0.963, and 0.963. These models achieved classifier performance measures of greater than 96%. In the future, XAI using agnostic models can be an additional important tool to help analyze data, including food products, even online. [ABSTRACT FROM AUTHOR]

Objective: To establish whether or not a natural language processing technique could identify two common inpatient neurosurgical comorbidities using only text reports of inpatient head imaging. Materials and methods: A training and testing dataset of reports of 979 CT or MRI scans of the brain for patients admitted to the neurosurgery service of a single hospital in June 2021 or to the Emergency Department between July 1–8, 2021, was identified. A variety of machine learning and deep learning algorithms utilizing natural language processing were trained on the training set (84% of the total cohort) and tested on the remaining images. A subset comparison cohort (n = 76) was then assessed to compare output of the best algorithm against real-life inpatient documentation. Results: For "brain compression", a random forest classifier outperformed other candidate algorithms with an accuracy of 0.81 and area under the curve of 0.90 in the testing dataset. For "brain edema", a random forest classifier again outperformed other candidate algorithms with an accuracy of 0.92 and AUC of 0.94 in the testing dataset. In the provider comparison dataset, for "brain compression," the random forest algorithm demonstrated better accuracy (0.76 vs 0.70) and sensitivity (0.73 vs 0.43) than provider documentation. For "brain edema," the algorithm again demonstrated better accuracy (0.92 vs 0.84) and AUC (0.45 vs 0.09) than provider documentation. Discussion: A natural language processing-based machine learning algorithm can reliably and reproducibly identify selected common neurosurgical comorbidities from radiology reports. Conclusion: This result may justify the use of machine learning-based decision support to augment provider documentation. [ABSTRACT FROM AUTHOR]

In this work, we used the convolutional neural network (CNN) method to invert sea surface height (SSH) from the Global Navigation Satellite System (GNSS) delayed Doppler map (DDM) data during 2009–2017 and compared the CNN inversion data with those obtained from traditional simple random forest (RF) method. SSH observations from the OSTM/Jason-2 satellite were used to judge the merits of the two methods. The results show that both methods yield good SSH inversion results, but when the training set is 9000, the root mean square errors of the SSH inversion results based on the CNN and the RF method are 16.78 and 15.96 respectively; as the training set increases above 9000, the accuracy of the CNN method is significantly better than that of the RF method. This suggests that SSH inversion based on the CNN method will become more advantageous as more data become available. [ABSTRACT FROM AUTHOR]

Invasive noxious weed species (INWS) are typical poisonous plants and forbs that are considered an increasing threat to the native alpine grassland ecosystems in the Qinghai–Tibetan Plateau (QTP). Accurate knowledge of the continuous cover of INWS across complex alpine grassland ecosystems over a large scale is required for their control and management. However, the cooccurrence of INWS and native grass species results in highly heterogeneous grass communities and generates mixed pixels detected by remote sensors, which causes uncertainty in classification. The continuous coverage of INWS at the pixel level has not yet been achieved. In this study, objective 1 was to test the capability of Senginel-2 imagery at estimating continuous INWS cover across complex alpine grasslands over a large scale and objective 2 was to assess the performance of the state-of-the-art convolutional neural network-based regression (CNNR) model in estimating continuous INWS cover. Therefore, a novel CNNR model and a random forest regression (RFR) model were evaluated for estimating INWS continuous cover using Sentinel-2 imagery. INWS continuous cover was estimated directly from Sentinel-2 imagery with an R2 ranging from 0.88 to 0.93 using the CNNR model. The RFR model combined with multiple features had a comparable accuracy, which was slightly lower than that of the CNNR model, with an R2 of approximately 0.85. Twelve green band-, red-edge band-, and near-infrared band-related features had important contributions to the RFR model. Our results demonstrate that the CNNR model performs well when estimating INWS continuous cover directly from Sentinel-2 imagery, and the RFR model combined with multiple features derived from the Sentinel-2 imager can also be used for INWS continuous cover mapping. Sentinel-2 imagery is suitable for mapping continuous INWS cover across complex alpine grasslands over a large scale. Our research provides information for the advanced mapping of the continuous cover of invasive species across complex grassland ecosystems or, more widely, terrestrial ecosystems over large spatial areas using remote sensors such as Sentinel-2. [ABSTRACT FROM AUTHOR]

Petrographic prediction is crucial in identifying target areas and understanding reservoir lithology in oil and gas exploration. Traditional logging methods often rely on manual interpretation and experiential judgment, which can introduce subjectivity and constraints due to data quality and geological variability. To enhance the precision and efficacy of lithology prediction, this study employed a Savitzky–Golay filter with a symmetric window for anomaly data processing, coupled with a residual temporal convolutional network (ResTCN) model tasked with completing missing logging data segments. A comparative analysis against the support vector regression and random forest regression model revealed that the ResTCN achieves the smallest MAE, at 0.030, and the highest coefficient of determination, at 0.716, which are indicative of its proximity to the ground truth. These methodologies significantly enhance the quality of the training data. Subsequently, a Transformer–long short-term memory (T-LS) model was applied to identify and classify the lithology of unexplored wells. The input layer of the Transformer model follows an embedding-like principle for data preprocessing, while the encoding block encompasses multi-head attention, Add & Norm, and feedforward components, integrating the multi-head attention mechanism. The output layer interfaces with the LSTM layer through dropout. A performance evaluation of the T-LS model against established rocky prediction techniques such as logistic regression, k-nearest neighbor, and random forest demonstrated its superior identification and classification capabilities. Specifically, the T-LS model achieved a precision of 0.88 and a recall of 0.89 across nine distinct lithology features. A Shapley analysis of the T-LS model underscored the utility of amalgamating multiple logging data sources for lithology classification predictions. This advancement partially addresses the challenges associated with imprecise predictions and limited generalization abilities inherent in traditional machine learning and deep learning models applied to lithology identification, and it also helps to optimize oil and gas exploration and development strategies and improve the efficiency of resource extraction. [ABSTRACT FROM AUTHOR]

The global spread of the SARS-CoV-2 pandemic, originating in Wuhan, China, has had profound consequences on both health and the economy. Traditional alignment-based phylogenetic tree methods for tracking epidemic dynamics demand substantial computational power due to the growing number of sequenced strains. Consequently, there is a pressing need for an alignment-free approach to characterize these strains and monitor the dynamics of various variants. In this work, we introduce a swift and straightforward tool named GenoSig, implemented in C++. The tool exploits the Di and Tri nucleotide frequency signatures to delineate the taxonomic lineages of SARS-CoV-2 by employing diverse machine learning (ML) and deep learning (DL) models. Our approach achieved a tenfold cross-validation accuracy of 87.88% (± 0.013) for DL and 86.37% (± 0.0009) for Random Forest (RF) model, surpassing the performance of other ML models. Validation using an additional unexposed dataset yielded comparable results. Despite variations in architectures between DL and RF, it was observed that later clades, specifically GRA, GRY, and GK, exhibited superior performance compared to earlier clades G and GH. As for the continental origin of the virus, both DL and RF models exhibited lower performance than in predicting clades. However, both models demonstrated relatively higher accuracy for Europe, North America, and South America compared to other continents, with DL outperforming RF. Both models consistently demonstrated a preference for cytosine and guanine over adenine and thymine in both clade and continental analyses, in both Di and Tri nucleotide frequencies signatures. Our findings suggest that GenoSig provides a straightforward approach to address taxonomic, epidemiological, and biological inquiries, utilizing a reductive method applicable not only to SARS-CoV-2 but also to similar research questions in an alignment-free context. [ABSTRACT FROM AUTHOR]

Aeroengines use numerous sensors to detect equipment health and ensure proper operation. Currently, filtering useful sensor data and removing useless data is challenging in predicting the remaining useful life (RUL) of an aeroengine using deep learning. To reduce computational costs and improve prediction performance, we use random forest to evaluate the feature importance of sensor data. Based on the size of the feature corresponding to the Gini index, we select the appropriate sensor. This helps us to determine which sensor to use and ensures that the computational resources are not wasted on unnecessary sensors. Considering that the RUL of equipment changes in a progressively more complex manner as the equipment is used over time, we propose an improved squeeze and excitation block (SSE) and combine it with a convolutional neural network (CNN). By enhancing the feature selection ability of CNN through segmented squeeze and excitation block, the model can focus on important information within features to effectively improve prediction performance. We compared our experiments with other RUL experiments on the CMAPSS aeroengine dataset and then conducted ablation experiments to verify the critical role of the methods we used. [ABSTRACT FROM AUTHOR]

At present, for business lightning positioning systems, the classification of lightning discharge types is mostly based on lightning pulse signal features, and there is still a lot of room for improvement. We propose a lightning discharge classification method based on convolutional encoding features. This method utilizes convolutional neural networks to extract encoding features, and uses random forests to classify the extracted encoding features, achieving high accuracy discrimination for various lightning discharge events. Compared with traditional multi-parameter-based methods, the new method proposed in this paper has the ability to identify multiple lightning discharge events and does not require precise detailed feature engineering to extract individual pulse parameters. The accuracy of this method for identifying lightning discharge types in intra-cloud flash (IC), cloud-to-ground flash (CG), and narrow bipolar events (NBEs) is 97%, which is higher than that of multi-parameter methods. Moreover, our method can complete the classification task of lightning signals at a faster speed. Under the same conditions, the new method only requires 28.2 µs to identify one pulse, while deep learning-based methods require 300 µs. This method has faster recognition speed and higher accuracy in identifying multiple discharge types, which can better meet the needs of real-time business positioning. [ABSTRACT FROM AUTHOR]

RNA-protein interaction (RPI) is crucial to the life processes of diverse organisms. Various researchers have identified RPI through long-term and high-cost biological experiments. Although numerous machine learning and deep learning-based methods for predicting RPI currently exist, their robustness and generalizability have significant room for improvement. This study proposes LPI-MFF, an RPI prediction model based on multi-source information fusion, to address these issues. The LPI-MFF employed protein–protein interactions features, sequence features, secondary structure features, and physical and chemical properties as the information sources with the corresponding coding scheme, followed by the random forest algorithm for feature screening. Finally, all information was combined and a classification method based on convolutional neural networks is used. The experimental results of fivefold cross-validation demonstrated that the accuracy of LPI-MFF on RPI1807 and NPInter was 97.60% and 97.67%, respectively. In addition, the accuracy rate on the independent test set RPI1168 was 84.9%, and the accuracy rate on the Mus musculus dataset was 90.91%. Accordingly, LPI-MFF demonstrated greater robustness and generalization than other prevalent RPI prediction methods. [ABSTRACT FROM AUTHOR]

In recent years, automatic image classification methods have significantly progressed, notably black box algorithms such as machine learning and deep learning. Unfortunately, such efforts only focused on improving performance, rather than attempting to explain and interpret how classification models actually operate. This article compares three state-of-the-art algorithms incorporating random forests, gradient boosting and convolutional neural networks for geomorphological mapping. It also attempts to explain how the most effective classifier makes decisions by evaluating which of the geomorphometric variables are most important for automatic mapping and how they affect the classification results using one of the explainable artificial intelligence techniques, namely accumulated local effects (ALE). This method allows us to understand the relationship between predictors and the model's outcome. For these purposes, eight sheets of the digital geomorphological map of Poland on the scale of 1:100,000 were used as the reference material. The classification results were validated using the holdout method and cross-validation for individual sheets representing different morphogenetic zones. The terrain elevation entropy, absolute elevation, aggregated median elevation and standard deviation of elevation had the greatest impact on the classification results among the 15 geomorphometric variables considered. The ALE analysis was conducted for the XGBoost classifier, which achieved the highest accuracy of 92.8%, ahead of Random Forests at 84% and LightGBM at 73.7% and U-Net at 59.8%. We conclude that automatic classification can support geomorphological mapping only if the geomorphological characteristics in the predicted area are similar to those in the training dataset. The ALE plots allow us to analyze the relationship between geomorphometric variables and landform membership, which helps clarify their role in the classification process. [ABSTRACT FROM AUTHOR]

The process of identifying apple varieties holds pivotal importance in pomology and agricultural science. This intricate task not only aids growers in optimizing orchard management, but also profoundly impacts consumers and the apple industry as a whole. Selecting the right apple varieties tailored to specific environmental conditions and market demands is instrumental for the sustainability and economic viability of apple cultivation. Accurate apple variety identification further contributes to maintaining product quality and ensuring consumer satisfaction. Traditional identification methods, however, are susceptible to human error given the vast diversity of apple cultivars. In response, the integration of advanced technologies, including image processing and machine learning, has emerged as a promising approach to enhance accuracy and efficiency in apple variety identification, benefitting both the agricultural and commercial sectors. The classification of apple types involved feature extraction using three methods: MobileNetV2, EfficientNetV2B0, and a combination of GLCM and Color-Space algorithms from apple images. Machine learning models were then built to classify apple varieties, utilizing various algorithms such as support vector machine (SVM), k-nearest neighbors (Knn), random subspace (RSS), and random forest. In the case of "EfficientNetV2B0 + GLCM + Color-Space" and utilizing the ReliefF feature selection method, the random forest algorithm attains peak performance with an accuracy, precision, recall, and F-score all registering an impressive 98.33%. [ABSTRACT FROM AUTHOR]

The research and application of convolutional neural networks (CNNs) on statistical downscaling have been hampered by the fact that deep learning is highly dependent on sample size and is considered to be a black-box model. Therefore, a CNN model with transfer learning (CNN-TL) is proposed to study the pre-rainy season precipitation of South China. First, an augmented monthly dataset is created by sliding a fixed-length window over the daily circulation field and precipitation data for the entire year. Next, a base CNN network is pretrained on the augmented dataset, and then the network parameters are tuned on the actual monthly dataset from South China. Then, guided backpropagation is conducted to obtain the distribution regions of the key features and explain the net. The coefficient of determination R2 and root-mean-square error (RMSE) show that the CNN-TL model has higher explanatory power and better fitting performance than the feature extraction–based random forest. In comparison with the base CNN, the transfer learning approach can improve the explanatory power of the model by 10.29% and reduce the average RMSE by 6.82%. In addition, the interpretation results of the model show that the critical regions are primarily South China and its surrounding areas, including the Indochina Peninsula, the Bay of Bengal, and the South China Sea. Furthermore, the ablation experiments and composite analysis illustrate that these regions are very important. Significance Statement: To mitigate the challenges posed by small sample sizes and the transparency of deep learning in downscaling problems, we propose a convolutional neural network based on sample augmentation and transfer learning to study the monthly precipitation downscaling problem during the preflood period in South China. In comparison with random forests and conventional convolutional neural networks, our model achieves an optimal interpretation rate and stability. In addition, we explore the interpretability of the model using guided backpropagation to find the distribution of key features within the large-scale circulation field, thus increasing the credibility of the model. [ABSTRACT FROM AUTHOR]

Airport service quality ratings found on social media such as Airline Quality and Google Maps offer invaluable insights for airport management to improve their quality of services. However, there is currently a lack of research analysing these reviews by airport services using sentimental analysis approaches. This research applies multiclass models based on Aspect-Based Sentimental Analysis to conduct a comprehensive analysis of travellers' reviews, in which the major airport services are tagged by positive, negative, and non-existent sentiments. Seven airport services commonly utilised in previous studies are also introduced. Subsequently, various Deep Learning architectures and Machine Learning classification algorithms are developed, tested, and compared using data collected from Twitter, Google Maps, and Airline Quality, encompassing travellers' feedback on airport service quality. The results show that the traditional Machine Learning algorithms such as the Random Forest algorithm outperform Deep Learning models in the multiclass prediction of airport service quality using travellers' feedback. The findings of this study offer concrete justifications for utilising multiclass Machine Learning models to understand the travellers' sentiments and therefore identify airport services required for improvement. [ABSTRACT FROM AUTHOR]

As the application of large language models (LLMs) has broadened into the realm of biological predictions, leveraging their capacity for self-supervised learning to create feature representations of amino acid sequences, these models have set a new benchmark in tackling downstream challenges, such as subcellular localization. However, previous studies have primarily focused on either the structural design of models or differing strategies for fine-tuning, largely overlooking investigations into the nature of the features derived from LLMs. In this research, we propose different ESM2 representation extraction strategies, considering both the character type and position within the ESM2 input sequence. Using model dimensionality reduction, predictive analysis and interpretability techniques, we have illuminated potential associations between diverse feature types and specific subcellular localizations. Particularly, the prediction of Mitochondrion and Golgi apparatus prefer segments feature closer to the N-terminal, and phosphorylation site-based features could mirror phosphorylation properties. We also evaluate the prediction performance and interpretability robustness of Random Forest and Deep Neural Networks with varied feature inputs. This work offers novel insights into maximizing LLMs' utility, understanding their mechanisms, and extracting biological domain knowledge. Furthermore, we have made the code, feature extraction API, and all relevant materials available at https://github.com/yujuan-zhang/feature-representation-for-LLMs. [ABSTRACT FROM AUTHOR]

Detecting road cracks is essential for inspecting and assessing the integrity of concrete pavement structures. Traditional image-based methods often require complex preprocessing to extract crack features, making them challenging when dealing with noisy concrete surfaces in diverse real-world scenarios, such as autonomous vehicle road detection. This study introduces an image-based crack detection approach that combines a Random Forest machine learning classifier with a deep convolutional neural network (CNN) to address these challenges. Three state-of-the-art models, namely MobileNet, InceptionV3, and Xception, were employed and trained using a dataset of 30,000 images to build an effective CNN. A systematic comparison of validation accuracy across various base learning rates identified a base learning rate of 0.001 as optimal, achieving a maximum validation accuracy of 99.97%. This optimal learning rate was then applied in the subsequent testing phase. The robustness and flexibility of the trained models were evaluated using 6,000 test photos, each with a resolution of 224 × 224 pixels, which were not part of the training or validation sets. The outstanding results, boasting a remarkable 99.95% accuracy, 99.95% precision, 99.94% recall, and a matching 99.94% F1 Score, unequivocally affirm the efficacy of the proposed technique in precisely identifying road fractures in photographs taken on real concrete surfaces. [ABSTRACT FROM AUTHOR]

A near-global framework for automated training data generation and land cover classification using shallow machine learning with low-density time series imagery does not exist. This study presents a methodology to map nine-class, six-class, and five-class land cover using two dates (winter and non-winter) of a Sentinel-2 granule across seven international sites. The approach uses a series of spectral, textural, and distance decision functions combined with modified ancillary layers (such as global impervious surface and global tree cover) to create binary masks from which to generate a balanced set of training data applied to a random forest classifier. For the land cover masks, stepwise threshold adjustments were applied to reflectance, spectral index values, and Euclidean distance layers, with 62 combinations evaluated. Global (all seven scenes) and regional (arid, tropics, and temperate) adaptive thresholds were computed. An annual 95th and 5th percentile NDVI composite was used to provide temporal corrections to the decision functions, and these corrections were compared against the original model. The accuracy assessment found that the regional adaptive thresholds for both the two-date land cover and the temporally corrected land cover could accurately map land cover type within nine-class (68.4% vs. 73.1%), six-class (79.8% vs. 82.8%), and five-class (80.1% vs. 85.1%) schemes. Lastly, the five-class and six-class models were compared with a manually labeled deep learning model (Esri), where they performed with similar accuracies (five classes: Esri 80.0 ± 3.4%, region corrected 85.1 ± 2.9%). The results highlight not only performance in line with an intensive deep learning approach, but also that reasonably accurate models can be created without a full annual time series of imagery. [ABSTRACT FROM AUTHOR]

Accurate extraction of crop acreage is an important element of digital agriculture. This study uses Sentinel-2A, Sentinel-1, and DEM as data sources to construct a multidimensional feature dataset encompassing spectral features, vegetation index, texture features, terrain features, and radar features. The Relief-F algorithm is applied for feature selection to identify the optimal feature dataset. And the combination of deep learning and the random forest (RF) classification method is utilized to identify lilies in Qilihe District and Yuzhong County of Lanzhou City, obtain their planting structure, and analyze their spatial distribution characteristics in Gansu Province. The findings indicate that terrain features significantly contribute to ground object classification, with the highest classification accuracy when the number of features in the feature dataset is 36. The precision of the deep learning classification method exceeds that of RF, with an overall classification accuracy and kappa coefficient of 95.9% and 0.934, respectively. The Lanzhou lily planting area is 137.24 km2, and it primarily presents a concentrated and contiguous distribution feature. The study's findings can serve as a solid scientific foundation for Lanzhou City's lily planting structure adjustment and optimization and a basis of data for local lily yield forecasting, development, and application. [ABSTRACT FROM AUTHOR]

Automated built-up infrastructure classification is a global need for planning. However, individual indices have weaknesses, including spectral confusion with bare ground, and computational requirements for deep learning are intensive. We present a computationally lightweight method to classify built-up infrastructure. We use an ensemble of spectral indices and a novel red-band texture layer with global thresholds determined from 12 diverse sites (two seasonally varied images per site). Multiple spectral indexes were evaluated using Sentinel-2 imagery. Our texture metric uses the red band to separate built-up infrastructure from spectrally similar bare ground. Our evaluation produced global thresholds by evaluating ground truth points against a range of site-specific optimal index thresholds across the 24 images. These were used to classify an ensemble, and then spectral indexes, texture, and stratified random sampling guided training data selection. The training data fit a random forest classifier to create final binary maps. Validation found an average overall accuracy of 79.95% (±4%) and an F1 score of 0.5304 (±0.07). The inclusion of the texture metric improved overall accuracy by 14–21%. A comparison to site-specific thresholds and a deep learning-derived layer is provided. This automated built-up infrastructure mapping framework requires only public imagery to support time-sensitive land management workflows. [ABSTRACT FROM AUTHOR]

Aquatic vegetation is an important component of aquatic ecosystems; therefore, the classification and mapping of aquatic vegetation is an important aspect of lake management. Currently, the decision tree (DT) classification method based on spectral indices has been widely used in the extraction of aquatic vegetation data, but the disadvantage of this method is that it is difficult to fix the threshold value, which, in turn, affects the automatic classification effect. In this study, Sentinel-2 MSI data were used to produce a sample set (about 930 samples) of aquatic vegetation in four inland lakes (Lake Taihu, Lake Caohai, Lake Honghu, and Lake Dongtinghu) using the visual interpretation method, including emergent, floating-leaved, and submerged vegetation. Based on this sample set, a DL model (Res-U-Net) was used to train an automatic aquatic vegetation extraction model. The DL model achieved a higher overall accuracy, relevant error, and kappa coefficient (90%, 8.18%, and 0.86, respectively) compared to the DT method (79%, 23.07%, and 0.77) and random forest (78%,10.62% and 0.77) when utilizing visual interpretation results as the ground truth. When utilizing measured point data as the ground truth, the DL model exhibited accuracies of 59%, 78%, and 91% for submerged, floating-leaved, and emergent vegetation, respectively. In addition, the model still maintains good recognition in the presence of clouds with the influence of water bloom. When applying the model to Lake Honghu from January 2017 to October 2023, the obtained temporal variation patterns in the aquatic vegetation were consistent with other studies. The study in this paper shows that the proposed DL model has good application potential for extracting aquatic vegetation data. [ABSTRACT FROM AUTHOR]

Abnormal changes in air temperature cause natural disasters such as droughts, hailstorms, and storms, thereby affecting the normal lives of human beings. Consequently, timely and accurate air temperature prediction is essential for human production and livelihood. Traditional air temperature prediction methods are less accurate and less consider the spatial relationship between air temperature in different regions. In this paper, we propose a new deep learning model, convolutional long short-term memory based on channel attention and spatial autocorrelation (ConvLSTM-CASA), which focuses on the spatial correlation between ambient air temperatures and can effectively capture the interaction of air temperatures in different regions. The results show that the ConvLSTM-CASA model has an average R2 of 0.954 and MSE of 5.245 for pentad-mean temperature prediction over the Yangtze River basin. Compared with baseline forecasting models, the MSE accuracy by the ConvLSTM-CASA model improved by 72.45%, 48.95%, 48.97%, 47.79%, and 22.63% over the decision tree regression (DTR), multiple linear regression (MLR), random forests (RF), long short-term memory (LSTM), and ConvLSTM models, respectively. The ConvLSTM-CASA model is expected to outperform the ConvLSTM model over 90% of the area, suggesting robust prediction skill improvement over space. The ConvLSTM-CASA model provides new insights for data-driven pentad-mean air temperature prediction by including elaborate channel and spatial feature modeling, which aid individuals in comprehending the intricate patterns of air temperature fluctuations. [ABSTRACT FROM AUTHOR]

This study investigates the prediction of mental well-being factors—depression, stress, and anxiety—using the NetHealth dataset from college students. The research addresses four key questions, exploring the impact of digital biomarkers on these factors, their alignment with conventional psychology literature, the time-based performance of applied methods, and potential enhancements through multitask learning. The findings reveal modality rankings aligned with psychology literature, validated against paper-based studies. Improved predictions are noted with temporal considerations, and further enhanced by multitasking. Mental health multitask prediction results show aligned baseline and multitask performances, with notable enhancements using temporal aspects, particularly with the random forest (RF) classifier. Multitask learning improves outcomes for depression and stress but not anxiety using RF and XGBoost. [ABSTRACT FROM AUTHOR]

Speech emotion recognition (SER) has gained an increased interest during the last decades as part of enriched affective computing. As a consequence, a variety of engineering approaches have been developed addressing the challenge of the SER problem, exploiting different features, learning algorithms, and datasets. In this paper, we propose the application of the graph theory for classifying emotionally-colored speech signals. Graph theory provides tools for extracting statistical as well as structural information from any time series. We propose to use the mentioned information as a novel feature set. Furthermore, we suggest setting a unique feature-based identity for each emotion belonging to each speaker. The emotion classification is performed by a Random Forest classifier in a Leave-One-Speaker-Out Cross Validation (LOSO-CV) scheme. The proposed method is compared with two state-of-the-art approaches involving well known hand-crafted features as well as deep learning architectures operating on mel-spectrograms. Experimental results on three datasets, EMODB (German, acted) and AESDD (Greek, acted), and DEMoS (Italian, in-the-wild), reveal that our proposed method outperforms the comparative methods in these datasets. Specifically, we observe an average UAR increase of almost 18 % , 8 % and 13 % , respectively. [ABSTRACT FROM AUTHOR]

In recent years, artificial intelligence (AI) has shown promising applications in various scientific domains, including biochemical analysis research. However, the effectiveness of AI in modeling small-scale, imbalanced datasets remains an open question in such fields. This study explores the capabilities of eight basic AI algorithms, including ridge regression, logistic regression, random forest regression, and others, in modeling a small, imbalanced clinical dataset (total n = 387, class 0 = 27, class 1 = 360) related to the records of the biochemical blood tests from the patients with multiple wasp stings (MWS). Through rigorous evaluation using k-fold cross-validation and comprehensive scoring, we found that none of the models could effectively model the data. Even after fine-tuning the hyperparameters of the best-performing models, the results remained below acceptable thresholds. The study highlights the challenges of applying AI to small-scale datasets with imbalanced groups in biochemical or clinical research and emphasizes the need for novel algorithms tailored to small-scale data. The findings also call for further exploration into techniques such as transfer learning and data augmentation, and they underline the importance of understanding the minimum dataset scale required for effective AI modeling in biochemical contexts. [ABSTRACT FROM AUTHOR]

This study presents a framework to attribute river runoff variations to the combined effects of reservoir operations, land surface changes, and climate variability. We delineated the data into natural and impacted periods. For the natural period, an integrated Long Short-Term Memory and Random Forest model was developed to accurately simulate both mean and extreme runoff values, outperforming existing models. This model was then used to estimate runoff unaffected by human activities in the impacted period. Our findings indicate stable annual and wet season mean runoff, with a decrease in wet season maximums and an increase in dry season means, while extreme values remained largely unchanged. A Budyko framework incorporating reconstructed runoff revealed that rainfall and land surface changes are the predominant factors influencing runoff variations in wet and dry seasons, respectively, and land surface impacts become more pronounced during the impacted period for both seasons. Human activities dominate dry season runoff variation (93.9%), with climate change at 6.1%, while in the wet season, the split is 64.5% to 35.5%. Climate change and human activities have spontaneously led to reduced runoff during the wet season and increased runoff during the dry season. Only reservoir regulation is found to be linked to human-induced runoff changes, while the effects of land surface changes remain ambiguous. These insights underscore the growing influence of anthropogenic factors on hydrological extremes and quantify the role of reservoirs within the impacts of human activities on runoff. [ABSTRACT FROM AUTHOR]

Kazakhstan is recently experiencing an increase in drought trends. However, low-capacity probabilistic drought forecasts and poor dissemination have led to a drought crisis in 2021 that resulted in the loss of thousands of livestock. To improve drought forecasting accuracy, this study applies Machine Learning and Deep Learning (ML and DL) algorithms to capture the sequences of drought events using a non-contiguous drought analysis (NCDA). Precipitation, 2-m temperature, runoff, solar radiation, relative humidity, and evaporation were collected from the ERA5 database as input variables. Combinations of inputs were used to build ML models, including seven classifiers (Logistic, K-NN, Kernel SVM, Decision Tree, Random Forest, XGBoost, and GRU). The output events were defined by standardized precipitation index (SPI) and SPEI indicators as binary classes. Weekly time series from 1991 to 2021 for each cell were used to forecast a lead time from 1 week to 6 months. GRU provided 97-99% accuracy in more volatile regions while Random Forest and XGBoost showed 94-99% accuracy at a lead time of 6 months. The accuracy evaluation was based on the confusion matrix and F1 score to analyze the stage change capture. This study demonstrates the effectiveness of using ML and DL algorithms for drought forecasting, with potential applications for other regions. [ABSTRACT FROM AUTHOR]

Photovoltaic (PV) power prediction plays a critical role amid the accelerating adoption of renewable energy sources. This paper introduces a bidirectional long short-term memory (BiLSTM) deep learning (DL) model designed for forecasting photovoltaic power one hour ahead. The dataset under examination originates from a small PV installation located at the Polytechnic School of the University of Alcala. To improve the quality of historical data and optimize model performance, a robust data preprocessing algorithm is implemented. The BiLSTM model is synergistically combined with a Bayesian optimization algorithm (BOA) to fine-tune its primary hyperparameters, thereby enhancing its predictive efficacy. The performance of the proposed model is evaluated across diverse meteorological and seasonal conditions. In deterministic forecasting, the findings indicate its superiority over alternative models employed in this research domain, specifically a multilayer perceptron (MLP) neural network model and a random forest (RF) ensemble model. Compared with the MLP and RF reference models, the proposed model achieves reductions in the normalized mean absolute error (nMAE) of 75.03% and 77.01%, respectively, demonstrating its effectiveness in this type of prediction. Moreover, interval prediction utilizing the bootstrap resampling method is conducted, with the acquired prediction intervals carefully adjusted to meet the desired confidence levels, thereby enhancing the robustness and flexibility of the predictions. [ABSTRACT FROM AUTHOR]

We present an evaluation of different deep learning and machine learning approaches for tree health classification in the Black Forest, the Harz Mountains, and the Göttinger Forest on a unique, highly accurate tree-level dataset. The multispectral UAV data were collected from eight forest plots with diverse tree species, mostly conifers. As ground truth data (GTD), nearly 1500 tree polygons with related attribute information on the health status of the trees were used. This data were collected during extensive fieldwork using a mobile application and subsequent individual tree segmentation. Extensive preprocessing included normalization, NDVI calculations, data augmentation to deal with the underrepresented classes, and splitting the data into training, validation, and test sets. We conducted several experiments using a classical machine learning approach (random forests), as well as different convolutional neural networks (CNNs)—ResNet50, ResNet101, VGG16, and Inception-v3—on different datasets and classes to evaluate the potential of these algorithms for tree health classification. Our first experiment was a binary classifier of healthy and damaged trees, which did not consider the degree of damage or tree species. The best results of a 0.99 test accuracy and an F1 score of 0.99 were obtained with ResNet50 on four band composites using the red, green, blue, and infrared bands (RGBI images), while VGG16 had the worst performance, with an F1 score of only 0.78. In a second experiment, we also distinguished between coniferous and deciduous trees. The F1 scores ranged from 0.62 to 0.99, with the highest results obtained using ResNet101 on derived vegetation indices using the red edge band of the camera (NDVIre images). Finally, in a third experiment, we aimed at evaluating the degree of damage: healthy, slightly damaged, and medium or heavily damaged trees. Again, ResNet101 had the best performance, this time on RGBI images with a test accuracy of 0.98 and an average F1 score of 0.97. These results highlight the potential of CNNs to handle high-resolution multispectral UAV data for the early detection of damaged trees when good training data are available. [ABSTRACT FROM AUTHOR]

In this work, we study the numerical solution of inverse eigenvalue problems from a machine learning perspective. Two different problems are considered: the inverse Sturm-Liouville eigenvalue problem for symmetric potentials and the inverse transmission eigenvalue problem for spherically symmetric refractive indices. Firstly, we solve the corresponding direct problems to produce the required eigenvalues datasets in order to train the machine learning algorithms. Next, we consider several examples of inverse problems and compare the performance of each model to predict the unknown potentials and refractive indices respectively, from a given small set of the lowest eigenvalues. The supervised regression models we use are k-Nearest Neighbours, Random Forests and Multi-Layer Perceptron. Our experiments show that these machine learning methods, under appropriate tuning on their parameters, can numerically solve the examined inverse eigenvalue problems. [ABSTRACT FROM AUTHOR]

Wheat is unquestionably the primary source of sustenance in human dietary intake. The cultivation areas and production capacity of wheat worldwide have been observed to increase in parallel with the growth of the global population. Wheat grains from different varieties, when mixed with durum wheat, result in a reduction in the protein content. Various types of wheat grains also exhibit the same characteristic. In this particular scenario, the significance of accurately categorizing wheat becomes more pronounced. In recent years, there has been a proliferation of studies aimed at categorizing agricultural products through the application of deep learning and machine learning methodologies. In the present study, a novel approach was introduced to simultaneously analyze deep features and image patches. This was achieved by utilizing a dataset consisting of a comprehensive collection of 8354 images, encompassing various bread wheat varieties. The classification of wheat types was carried out by employing feature extraction using three distinct methods. MobileNetV2, EfficientNetV2B0, GLCM, and Color-Space algorithms were employed to extract features from the images. Lastly, the Support Vector Machine (SVM), Random Subspace ensemble with k-Nearest Neighbors (RSeslibKnn), Artificial Neural Network (ANN), and Random Forest algorithms were employed to develop models for the classification of bread wheat images. The evaluation of the experimental performances was conducted based on the criteria of accuracy, precision, recall, F-score, and mean absolute error (MAE). In general, the obtained accuracies ranged from 91.50 to 98.65%, which demonstrates the models' proficiency in accurately classifying the samples. When examining different algorithms, Support Vector Machines (SVM) consistently demonstrate robust performance by achieving high levels of accuracy, precision, recall, and F-scores across various feature combinations. [ABSTRACT FROM AUTHOR]

Throughout evolution, pathogenic viruses have developed different strategies to evade the response of the adaptive immune system. To carry out successful replication, some pathogenic viruses encode different proteins that manipulate the molecular mechanisms of host cells. Currently, there are different bioinformatics tools for virus research; however, none of them focus on predicting viral proteins that evade the adaptive system. In this work, we have developed a novel tool based on machine and deep learning for predicting this type of viral protein named VirusHound-I. This tool is based on a model developed with the multilayer perceptron algorithm using the dipeptide composition molecular descriptor. In this study, we have also demonstrated the robustness of our strategy for data augmentation of the positive dataset based on generative antagonistic networks. During the 10-fold cross-validation step in the training dataset, the predictive model showed 0.947 accuracy, 0.994 precision, 0.943 F1 score, 0.995 specificity, 0.896 sensitivity, 0.894 kappa, 0.898 Matthew's correlation coefficient and 0.989 AUC. On the other hand, during the testing step, the model showed 0.964 accuracy, 1.0 precision, 0.967 F1 score, 1.0 specificity, 0.936 sensitivity, 0.929 kappa, 0.931 Matthew's correlation coefficient and 1.0 AUC. Taking this model into account, we have developed a tool called VirusHound-I that makes it possible to predict viral proteins that evade the host's adaptive immune system. We believe that VirusHound-I can be very useful in accelerating studies on the molecular mechanisms of evasion of pathogenic viruses, as well as in the discovery of therapeutic targets. [ABSTRACT FROM AUTHOR]

Citrus is an important commercial crop in many areas. The management and planning of citrus growing can be supported by timely and efficient monitoring of citrus-growing regions. Their complex planting structure and the weather are likely to cause problems for extracting citrus-growing regions from remote sensing images. To accurately extract citrus-growing regions, deep learning is employed, because it has a strong feature representation ability and can obtain rich semantic information. A novel model for extracting citrus-growing regions by UNet that incorporates an image pyramid structure is proposed on the basis of the Sentinel-2 satellite imagery. A pyramid-structured encoder, a decoder, and multiscale skip connections are the three main components of the model. Additionally, atrous spatial pyramid pooling is used to prevent information loss and improve the ability to learn spatial features. The experimental results show that the proposed model has the best performance, with the precision, the intersection over union, the recall, and the F1-score reaching 88.96%, 73.22%, 80.55%, and 84.54%, respectively. The extracted citrus-growing regions have regular boundaries and complete parcels. Furthermore, the proposed model has greater overall accuracy, kappa, producer accuracy, and user accuracy than the object-oriented random forest algorithm that is widely applied in various fields. Overall, the proposed method shows a better generalization ability, higher robustness, greater accuracy, and less fragmented extraction results. This research can support the rapid and accurate mapping of large-scale citrus-growing regions. [ABSTRACT FROM AUTHOR]

The total electron content (TEC) is a crucial parameter for ionosphere monitoring, and the development of accurate TEC estimation and prediction models is significant for high-precision positioning. However, establishing a high-precision ionospheric model that can effectively improve prediction accuracy remains a challenging task. We propose a method that combines random forest with a Bi-LSTM neural network to establish a high-precision ionospheric prediction model. The random forest algorithm is used for regression analysis and to estimate the variable importance of input parameters. We use observation data from the Crustal Movement Observation Network of China and International GNSS Services while estimating the relative importance of 14 input variables at different latitudes. The results demonstrate that our proposed model achieves more efficient and accurate predictions, with a correlation coefficient of 0.96, indicating significant improvement in accuracy compared to traditional RNN and LSTM models. Overall, the Bi-LSTM forecast model based on random forest can effectively capture the temporal and spatial variation characteristics of ionospheric TEC in China. This enables the model to provide accurate ionospheric information for positioning users, thereby improving the precision of positioning applications. [ABSTRACT FROM AUTHOR]

The development of renewable resources and the deregulation of the market have made forecasting energy demand more critical in recent years. Advanced intelligent models are created to ensure accurate power projections for several time horizons to address new difficulties. Intelligent forecasting algorithms are a fundamental component of smart grids and a powerful tool for reducing uncertainty in order to make more cost- and energy-efficient decisions about generation scheduling, system reliability and power optimization, and profitable smart grid operations. However, since many crucial tasks of power operators, such as load dispatch, rely on short-term forecasts, prediction accuracy in forecasting algorithms is highly desired. This essay suggests a model for estimating Denmark’s power use that can precisely forecast the month’s demand. In order to identify factors that may have an impact on the pattern of a number of unique qualities in the city direct consumption of electricity. The current paper also demonstrates how to use an ensemble deep learning technique and Random forest to dramatically increase prediction accuracy. In addition to their ensemble, we showed how well the individual Random forest performed. [ABSTRACT FROM AUTHOR]

Speech is one of the most serious manifestations of Parkinson's disease (PD). Sophisticated language/speech models have already demonstrated impressive performance on a variety of tasks, including classification. By analysing large amounts of data from a given setting, these models can identify patterns that would be difficult for clinicians to detect. We focus on evaluating the performance of a large self-supervised speech representation model, wav2vec, for PD classification. Based on the computed wav2vec embedding for each available speech signal, we calculated two sets of 512 derived features, wav2vec-sum and wav2vec-mean. Unlike traditional signal processing methods, this approach can learn a suitable representation of the signal directly from the data without requiring manual or hand-crafted feature extraction. Using an ensemble random forest classifier, we evaluated the embedding-based features on three different healthy vs. PD datasets (participants rhythmically repeat syllables /pa/, Italian dataset and English dataset). The obtained results showed that the wav2vec signal representation was accurate, with a minimum area under the receiver operating characteristic curve (AUROC) of 0.77 for the /pa/ task and the best AUROC of 0.98 for the Italian speech classification. The findings highlight the potential of the generalisability of the wav2vec features and the performance of these features in the cross-database scenarios. [ABSTRACT FROM AUTHOR]

Ensuring food security in precision agriculture requires early prediction of soybean yield at various scales within the United States (U.S.), ranging from international to local levels. Accurate yield estimation is essential in preventing famine by providing insights into food availability during the growth season. Numerous deep learning (DL) algorithms have been developed to estimate soybean yield effectively using time-series remote sensing (RS) data to achieve this goal. However, the training data with short time spans can limit their ability to adapt to the dynamic and nuanced temporal changes in crop conditions. To address this challenge, we designed a 3D-ResNet-BiLSTM model to efficiently predict soybean yield at the county level across the U.S., even when using training data with shorter periods. We leveraged detailed Sentinel-2 imagery and Sentinel-1 SAR images to extract spectral bands, key vegetation indices (VIs), and VV and VH polarizations. Additionally, Daymet data was incorporated via Google Earth Engine (GEE) to enhance the model's input features. To process these inputs effectively, a dedicated 3D-ResNet architecture was designed to extract high-level features. These enriched features were then fed into a BiLSTM layer, enabling accurate prediction of soybean yield. To evaluate the efficacy of our model, its performance was compared with that of well-known models, including the Linear Regression (LR), Random Forest (RF), and 1D/2D/3D-ResNet models, as well as a 2D-CNN-LSTM model. The data from a short period (2019 to 2020) were used to train all models, while their accuracy was assessed using data from the year 2021. The experimental results showed that the proposed 3D-Resnet-BiLSTM model had a superior performance compared to the other models, achieving remarkable metrics (R2 = 0.791, RMSE = 5.56 Bu Ac−1, MAE = 4.35 Bu Ac−1, MAPE = 9%, and RRMSE = 10.49%). Furthermore, the 3D-ResNet-BiLSTM model showed a 7% higher R2 than the ResNet and RF models and an enhancement of 27% and 17% against the LR and 2D-CNN-LSTM models, respectively. The results highlighted our model's potential for accurate soybean yield predictions, supporting sustainable agriculture and food security. [ABSTRACT FROM AUTHOR]

Puerariae Thomsonii Radix (PTR) is not only widely used in disease prevention and treatment but is also an important raw material as a source of starch and other food. The growth years of PTR are closely related to its quality. The rapid and nondestructive identification of growth year is essential for the quality control of PTR and other traditional Chinese medicines. In this study, we proposed a convolutional neural network (CNN)-based classification framework in conjunction with hyperspectral imaging (HSI) technology for the rapid identification of the growth years of PTRs. Traditional treatment methods (i.e., multiplicative scatter correction, standard normal variate, and Savitzky-Golay smoothing) combined with machine learning algorithms (i.e., random forest, logistic regression, naive Bayes, and eXtreme gradient boost) were used as baseline models. Among them, the F1-score of CNN-based models based on PTRs' outer surfaces was over 90%, outperforming all the other baseline models. These results showed that it was feasible to use a deep learning algorithm in conjunction with HSI technology to identify the growth years of PTR. This method provides a fast, nondestructive, and simple method of identifying the growth years of PTR. It can be easily applied to other scenarios, such as for the identification of the locality or years of growth for other traditional Chinese herbs. [ABSTRACT FROM AUTHOR]

Many recently proposed structure-based virtual screening models appear to be able to accurately distinguish high affinity binders from non-binders. However, several recent studies have shown that they often do so by exploiting ligand-specific biases in the dataset, rather than identifying favourable intermolecular interactions in the input protein-ligand complex. In this work we propose a novel approach for assessing the extent to which machine learning-based virtual screening models are able to identify the functional groups responsible for binding. To sidestep the difficulty in establishing the ground truth importance of each atom of a large scale set of protein-ligand complexes, we propose a protocol for generating synthetic data. Each ligand in the dataset is surrounded by a randomly sampled point cloud of pharmacophores, and the label assigned to the synthetic protein-ligand complex is determined by a 3-dimensional deterministic binding rule. This allows us to precisely quantify the ground truth importance of each atom and compare it to the model generated attributions. Using our generated datasets, we demonstrate that a recently proposed deep learning-based virtual screening model, PointVS, identified the most important functional groups with 39% more efficiency than a fingerprint-based random forest, suggesting that it would generalise more effectively to new examples. In addition, we found that ligand-specific biases, such as those present in widely used virtual screening datasets, substantially impaired the ability of all ML models to identify the most important functional groups. We have made our synthetic data generation framework available to facilitate the benchmarking of new virtual screening models. Code is available at https://github.com/tomhadfield95/synthVS. [ABSTRACT FROM AUTHOR]

Accurately grasping the distribution and area of cotton for agricultural irrigation scheduling, intensive and efficient management of water resources, and yield estimation in arid and semiarid regions is of great significance. In this paper, taking the Xinjiang Shihezi oasis agriculture region as the study area, extracting the spectroscopic characterization (R, G, B, panchromatic), texture feature (entropy, mean, variance, contrast, homogeneity, angular second moment, correlation, and dissimilarity) and characteristics of vegetation index (normalized difference vegetation index/NDVI, ratio vegetation index/DVI, difference vegetation index/RVI) in the cotton flowering period before and after based on GF-6 image data, four models such as the random forests (RF) and deep learning approach (U-Net, DeepLabV3+ network, Deeplabv3+ model based on attention mechanism) were used to identify cotton and to compare their accuracies. The results show that the deep learning model is better than that of the random forest model. In all the deep learning models with three kinds of feature sets, the recognition accuracy and credibility of the DeepLabV3+ model based on the attention mechanism are the highest, the overall recognition accuracy of cotton is 98.23%, and the kappa coefficient is 96.11. Using the same Deeplabv3+ model based on an attention mechanism with different input feature sets (all features and only spectroscopic characterization), the identification accuracy of the former is much higher than that of the latter. GF-6 satellite image data in the field of crop type recognition has great application potential and prospects. [ABSTRACT FROM AUTHOR]

Wearable devices have become ubiquitous, collecting rich temporal data that offers valuable insights into human activities, health monitoring, and behavior analysis. Leveraging these data, researchers have developed innovative approaches to classify and predict time-based patterns and events in human life. Time-based techniques allow the capture of intricate temporal dependencies, which is the nature of the data coming from wearable devices. This paper focuses on predicting well-being factors, such as stress, anxiety, and positive and negative affect, on the Tesserae dataset collected from office workers. We examine the performance of different methodologies, including deep-learning architectures, LSTM, ensemble techniques, Random Forest (RF), and XGBoost, and compare their performances for time-based and non-time-based versions. In time-based versions, we investigate the effect of previous records of well-being factors on the upcoming ones. The overall results show that time-based LSTM performs the best among conventional (non-time-based) RF, XGBoost, and LSTM. The performance even increases when we consider a more extended previous period, in this case, 3 past-days rather than 1 past-day to predict the next day. Furthermore, we explore the corresponding biomarkers for each well-being factor using feature ranking. The obtained rankings are compatible with the psychological literature. In this work, we validated them based on device measurements rather than subjective survey responses. [ABSTRACT FROM AUTHOR]

Efficient land management and farming practices are critical to maintaining agricultural production, especially in Europe with limited arable land. It is very time consuming to rely on a manual field inspection of cultivated land to archive farm crops. But with the help of satellite monitoring data on the earth's surface, it is a new vision to classify farmland based on deep learning. This article has studied the Sentinel 2 (S2) data, which are top-of-atmosphere (TOA) reflectance values at the processing level-1C (L1C) observed from some areas of Germany and France. Aiming at the problem that the interference of atmosphere and cloud coverage weakens the recognition accuracy of subsequent algorithms, a method of combining feature expansion and feature importance analysis is proposed to optimize the raw S2 data. Specifically, the new 13 spectral features are expanded based on the linear and nonlinear combination of the raw 13 spectral bands of S2. The random forest (RF) algorithm is used to score the importance of features, and the important features of each time series are selected to form a new dataset. Then, an end-to-end deep learning model has been used for training. The structure of the model is a two-layer unidirectional recurrent neural network with long short-term memory (LSTM) as the backbone. And two linear layers as the output, which form two decision-making heads, respectively, representing output classification probability and the stop decision. The results show that adding features and selecting features is beneficial for the model to improve classification accuracy and predict the classification without all of the input data. This end-to-end classification pattern with early prediction would support intelligent monitoring of farm crops with a great advantage to the implementation of various agricultural policies. [ABSTRACT FROM AUTHOR]

Computer-aided diagnosis techniques based on deep learning in skin cancer classification have disadvantages such as unbalanced datasets, redundant information in the extracted features and ignored interactions of partial features among different convolutional layers. In order to overcome these disadvantages, we propose a skin cancer classification model named EFFNet, which is based on feature fusion and random forests. Firstly, the model preprocesses the HAM10000 dataset to make each category of training set images balanced by image enhancement technology. Then, the pre-training weights of the EfficientNetV2 model on the ImageNet dataset are fine-tuned on the HAM10000 skin cancer dataset. After that, an improved hierarchical bilinear pooling is introduced to capture the interactions of some features between the layers and enhance the expressive ability of features. Finally, the fused features are passed into the random forests for classification prediction. The experimental results show that the accuracy, recall, precision and F1-score of the model reach 94.96%, 93.74%, 93.16% and 93.24% respectively. Compared with other models, the accuracy rate is improved to some extent and the highest accuracy rate can be increased by about 10%. [ABSTRACT FROM AUTHOR]

Present approaches in PV (Photovoltaic) detection are known to be scalable to a larger area using machine learning classification and have improved accuracy on a regional scale with deep learning diagnostics. However, it may cause false detection, time, and cost-consuming when regional deep learning models are directly scaled to a larger area, particularly in large-scale, highly urbanized areas. Thus, a novel two-step downscaling methodology integrating machine learning broad spatial partitioning (step-1) and detailed deep learning diagnostics (step-2) is designed and applied in highly urbanized Jiangsu Province, China. In the first step, this methodology selects suitable feature combinations using the recursive feature elimination with distance correlation coefficient (RFEDCC) strategy for the random forest (RF), considering not only feature importance but also feature independence. The results from RF (overall accuracy = 95.52%, Kappa = 0.91) indicate clear boundaries and little noise. Furthermore, the post-processing of noise removal with a morphological opening operation for the extraction result of RF is necessary for the purpose that less high-resolution remote sensing tiles should be applied in the second step. In the second step, tiles intersecting with the results of the first step are selected from a vast collection of Google Earth tiles, reducing the computational complexity of the next step in deep learning. Then, the improved HRNet with high performance on the test data set (Intersection over Union around 94.08%) is used to extract PV plants from the selected tiles, and the results are mapped. In general, for Jiangsu province, the detection rate of the previous PV database is higher than 92%, and this methodology reduces false detection noise and time consumption (around 95%) compared with a direct deep learning methodology. [ABSTRACT FROM AUTHOR]

There is a critical need for a global monitoring capability for Tailings Storage Facilities (TSFs), to help protect the surrounding communities and the environment. Satellite Synthetic Aperture Radar Interferometry (InSAR) shows much promise towards this ambition. However, extracting meaningful information and interpreting the deformation patterns from InSAR data can be a challenging task. One approach to address this challenge is through the use of data science techniques. In this study, the representation of InSAR metadata as Entity Embeddings within a Deep Learning framework (EE-DL) is investigated for modelling the spatio-temporal deformation response. Entity embeddings are commonly used in natural-language-processing tasks. They represent discrete objects, such as words, as continuous, low-dimensional vectors that can be manipulated mathematically. We demonstrate that EE-DL can be used to predict anomalous patterns in the InSAR time series. To evaluate the performance of the EE-DL approach in SAR interferometry, we conducted experiments over a mining test site (Cadia, Australia), which has been subject to a TSF failure. This study demonstrated that EE-DL can detect and predict the fine spatial movement patterns that eventually resulted in the failure. We also compared the results with deformation predictions from common baseline models, the Random Forest model and Gaussian Process Regression (GPR). Both EE-DL and GPR greatly outperform Random Forest. While GPR is also able to predict displacement patterns with millimetric accuracy, it detects a significantly lower number of anomalies compared to EE-DL. Overall, our study showed that EE-DL is a promising approach for building early-warning systems for critical infrastructures that use InSAR to predict ground deformations. [ABSTRACT FROM AUTHOR]

The analysis and recognition of sign languages are currently active fields of research focused on sign recognition. Various approaches differ in terms of analysis methods and the devices used for sign acquisition. Traditional methods rely on video analysis or spatial positioning data calculated using motion capture tools. In contrast to these conventional recognition and classification approaches, electromyogram (EMG) signals, which measure muscle electrical activity, offer potential technology for detecting gestures. These EMG-based approaches have recently gained attention due to their advantages. This prompted us to conduct a comprehensive study on the methods, approaches, and projects utilizing EMG sensors for sign language handshape recognition. In this paper, we provided an overview of the sign language recognition field through a literature review, with the objective of offering an in-depth review of the most significant techniques. These techniques were categorized in this article based on their respective methodologies. The survey discussed the progress and challenges in sign language recognition systems based on surface electromyography (sEMG) signals. These systems have shown promise but face issues like sEMG data variability and sensor placement. Multiple sensors enhance reliability and accuracy. Machine learning, including deep learning, is used to address these challenges. Common classifiers in sEMG-based sign language recognition include SVM, ANN, CNN, KNN, HMM, and LSTM. While SVM and ANN are widely used, random forest and KNN have shown better performance in some cases. A multilayer perceptron neural network achieved perfect accuracy in one study. CNN, often paired with LSTM, ranks as the third most popular classifier and can achieve exceptional accuracy, reaching up to 99.6% when utilizing both EMG and IMU data. LSTM is highly regarded for handling sequential dependencies in EMG signals, making it a critical component of sign language recognition systems. In summary, the survey highlights the prevalence of SVM and ANN classifiers but also suggests the effectiveness of alternative classifiers like random forests and KNNs. LSTM emerges as the most suitable algorithm for capturing sequential dependencies and improving gesture recognition in EMG-based sign language recognition systems. [ABSTRACT FROM AUTHOR]