Music Genre Classification with LSTM based on Time and Frequency Domain Features

Music genre classification has become a big task in the age of seamless music streaming services and content development. Accurate music genre classification is critical for applications such as music recommendations, content organization and identifying musical trends. This study proposes a comprehensive approach to music genre classification based on deep learning and advanced audio analysis tools. To address the challenge of accurately classifying music genres for music apps, a deep learning model is used with the GTZAN dataset for classification. The study investigates on various algorithms like Convolutional Neural Networks (CNN), Feedforward Neural Networks (FNN), Support Vector Machines (SVM), k-Nearest Neighbor (kNN) and Long Short-term Memory (LSTM) on this dataset. The study involves cross-validating the models output after feature extraction from the preprocessed audio data and evaluating its performance. This paper proposes an enhanced CNN model which outperforms regular neural networks by taking advantage of its capacity to record complex spectral patterns. These findings show how deep learning algorithms can improve systems for categorizing music genres, potentially benefiting music-related apps and interfaces. Hence, the proposed model has given an accuracy of 94.2% on the GTZAN dataset and 93.3% on the ISMIR2004 Ballroom dataset.

PurposePenetration state is one of the most important factors for judging the quality of a gas tungsten arc welding (GTAW) joint. The purpose of this paper is to identify and classify the penetration state and welding quality through the features of arc sound signal during robotic GTAW process.Design/methodology/approachThis paper tried to make a foundation work to achieve on‐line monitoring of penetration state to weld pool through arc sound signal. The statistic features of arc sound under different penetration states like partial penetration, full penetration and excessive penetration were extracted and analysed, and wavelet packet analysis was used to extract frequency energy at different frequency bands. The prediction models were established by artificial neural networks based on different features combination.FindingsThe experiment results demonstrated that each feature in time and frequency domain could react the penetration behaviour, arc sound in different frequency band had different performance at different penetration states and the prediction model established by 23 features in time domain and frequency domain got the best prediction effect to recognize different penetration states and welding quality through arc sound signal.Originality/valueThis paper tried to make a foundation work to achieve identifying penetration state and welding quality through the features of arc sound signal during robotic GTAW process. A total of 23 features in time domain and frequency domain were extracted at different penetration states. And energy at different frequency bands was proved to be an effective factor for identifying different penetration states. Finally, a prediction model built by 23 features was proved to have the best prediction effect of welding quality.

Many countries around the globe, including Tanzania, are investing in renewable energy, partly because of the need to increase electric power generation and meet the ever-increasing electricity demands, and partly because of the need to combat global climate change caused by carbon emissions from fossil fuel-based electricity production. Kishapu Solar Project, located in Kishapu District, Shinyanga Region, Tanzania, is one of the renewable energy projects with a planned generation of 150 MW of electric power once completed, using solar photovoltaic technology. Accurate forecasting of solar radiation is important in managing and maintaining solar energy plants. Inaccurate forecasting of solar radiation can lead to poor harvest of solar energy and poor management of solar energy plants, which directly impacts electricity contribution to the national electricity grid. Long Short-Term Memory (LSTM) and its variant, PeepHole LSTM, are examples of Deep Learning models that can be used to predict solar radiation. The purpose of PeepHole LSTM is to introduce “peephole” connections in the LSTM model, which allow LSTM gates to directly inspect memory cell state and hence have richer information to process. However, there is still a research gap on the impact of “peephole” connections on the prediction performance of the LSTM model, especially in the context of forecasting solar radiation in Kishapu District, where the Kishapu Solar Project is located. The purpose of this study is to fill this research gap by developing both LSTM and PeepHole LSTM models and comparing their performances in forecasting solar radiation in Kishapu District. The 14-year Kishapu District daily solar radiation (Downward Shortwave Radiation) dataset was first pre-processed by scaling (normalizing) it using Standard Scaler and then split into training set (the first 70%), validation set (the next 15%) and test set (the last 15%) before being utilized to train and test the developed LSTM and PeepHole LSTM models. Both LSTM and PeepHole LSTM models used the previous 5 days' solar radiation values as inputs to predict the output (next day (6th day) solar radiation value). The trained and better-performing model between the two was finally saved in .h5 format and integrated with a Gradio based Web Application to provide a user interface for officials in Kishapu District and at Kishapu Solar Project to predict daily solar radiation. The findings from performance evaluation revealed that the LSTM model outperformed the PeepHole LSTM model in predicting solar radiation by achieving test Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) scores of 26.22 W/m2 and 36.65 W/m2, respectively, compared to MAE and RMSE scores of 30.84 W/m2 and 41.88 W/m2, respectively, achieved by the PeepHole LSTM model, concluding that “peephole” connections reduce the performance of the LSTM model rather than increasing it when used to forecast solar radiation in Kishapu District. This study’s recommendation to Artificial Intelligence (AI) software developers and researchers is to use the LSTM rather than the PeepHole LSTM Deep Learning model when forecasting solar radiation in Kishapu District and in locations with similar climatic conditions

With the increasing awareness of health, using wearable sensors to monitor individual activities and accurately estimate energy expenditure has become a current research focus. However, existing research encounters challenges including low estimation accuracy, a deficiency of frequency domain features, and difficulty in integrating time domain and frequency domain features. To address these issues, we propose an innovative framework called the Dual-Stream Fusion Network (DSFN). This framework combines the Time Domain Encoding (TDE) module, the Frequency Domain Hierarchical-Split Encoding (FDHSE) module, and a Two-Stage Feature Fusion (TSF) module. Specifically, the temporal stream of the framework employs the TDE module to capture deep temporal features that reflect the complex dynamic variations in time-series data. The frequency domain stream introduces the FDHSE module, which extracts frequency domain features using a multi-level, multi-scale approach, ensuring a comprehensive and diverse representation of frequency information. Through this dual-stream architecture, our model effectively learns both time and frequency domain features, addressing the limitations of frequency domain features observed in prior studies. Additionally, we propose the TSF module to fully integrate time and frequency domain features, effectively overcoming the challenge of fusing these two types of features. We conducted experiments on two public datasets, namely the GOTOV dataset (elderly people) and the JSI dataset (young people). Experimental results demonstrate that our method achieves excellent performance across different age groups. Compared to the baseline models, the proposed DSFN significantly improves the accuracy of human energy expenditure estimation.

BackgroundPredicting hypoglycemia while maintaining a low false alarm rate is a challenge for the wide adoption of continuous glucose monitoring (CGM) devices in diabetes management. One small study suggested that a deep learning model based on the long short-term memory (LSTM) network had better performance in hypoglycemia prediction than traditional machine learning algorithms in European patients with type 1 diabetes. However, given that many well-recognized deep learning models perform poorly outside the training setting, it remains unclear whether the LSTM model could be generalized to different populations or patients with other diabetes subtypes.ObjectiveThe aim of this study was to validate LSTM hypoglycemia prediction models in more diverse populations and across a wide spectrum of patients with different subtypes of diabetes.MethodsWe assembled two large data sets of patients with type 1 and type 2 diabetes. The primary data set including CGM data from 192 Chinese patients with diabetes was used to develop the LSTM, support vector machine (SVM), and random forest (RF) models for hypoglycemia prediction with a prediction horizon of 30 minutes. Hypoglycemia was categorized into mild (glucose=54-70 mg/dL) and severe (glucose<54 mg/dL) levels. The validation data set of 427 patients of European-American ancestry in the United States was used to validate the models and examine their generalizations. The predictive performance of the models was evaluated according to the sensitivity, specificity, and area under the receiver operating characteristic curve (AUC).ResultsFor the difficult-to-predict mild hypoglycemia events, the LSTM model consistently achieved AUC values greater than 97% in the primary data set, with a less than 3% AUC reduction in the validation data set, indicating that the model was robust and generalizable across populations. AUC values above 93% were also achieved when the LSTM model was applied to both type 1 and type 2 diabetes in the validation data set, further strengthening the generalizability of the model. Under different satisfactory levels of sensitivity for mild and severe hypoglycemia prediction, the LSTM model achieved higher specificity than the SVM and RF models, thereby reducing false alarms.ConclusionsOur results demonstrate that the LSTM model is robust for hypoglycemia prediction and is generalizable across populations or diabetes subtypes. Given its additional advantage of false-alarm reduction, the LSTM model is a strong candidate to be widely implemented in future CGM devices for hypoglycemia prediction.

Objectives: To facilitate painless and easy method for prediction of diabetes with high accuracy and to measure blood glucose by noninvasive method using Photoplethysmography (PPG). Method: In this study, diabetes prediction is done using different machine learning algorithms on a dataset created by using samples from PIMA Indian Diabetes dataset and in vivo diabetes dataset. Machine learning algorithms used are Support Vector Machine (SVM), Decision Tree, Naïve Bayes Classifier and K Nearest Neighbor (KNN). A PPG data of 182 individuals is recorded over 1 minute duration each. Various frequency and time domain features of PPG single are extracted using single pulse wave analysis. Neural network is trained using extracted features and glucose measurement is performed. Findings: With decision tree algorithm we got highest accuracy of 89.97% for diabetes prediction and it proves to be good algorithm for consideration for measurement in diabetes treatment.Clarke Error Grid analysis for blood glucose prediction is clinically accepted, so we performed similar analysis. Using time and frequency domain features, we got 94.27 % data points in clinically accepted regions (Region A and Region B). Novelty: Based on the data collected or samples analyzed, and accuracy of our results, it is encouraging to see that further research may lead to affordable noninvasive method for detection of diabetes at early stage. Keywords: Noninvasive; blood glucose; NIR; machine learning algorithms; photoplethysmography

In the study of detection of an epileptic seizure using Electroencephalogram (EEG), pattern recognition has been recognized as a valued tool. In this pattern recognition study, the first time the authors have attempted to use time domain (TD) features such as waveform length (WL), number of zero-crossings (ZC) and number of slope sign changes (SSC) derived directly from filtered EEG data and from discrete wavelet transform (DWT) of filtered EEG data for the detection of an epileptic seizure. Further, the authors attempted to study the performance of other time domain features such as mean absolute value (MAV), standard deviation (SD), average power (AVP), which had been attempted by other researchers. The performance of the TD features is studied using naive Bayes (NB) and support vector machines (SVM) classifiers for the university of Bonn database with fourteen different combinations of set E with set A to D. The proposed scheme was also compared with other existing scheme in the literature. The implementation results showed that the proposed scheme could attain the highest accuracy of 100% for normal eyes open and epileptic data set with direct as well as DWT based TD features. For other data sets, the highest accuracy is obtained with DWT based TD features using SVM. However, no significant difference in the classification of 14 data sets with TD features filtered EEG data and from DWT of filtered EEG data.

Wearable electroencephalography (EEG) devices for seizure detection accuracy and reliability are deep learning (DL) applications in the field of epilepsy diagnosis. In this study, we sought to increase the accuracy of seizure detection using advanced DL algorithms on the Children’s Hospital Boston - Massachusetts Institute of Technology (CHB-MIT) EEG database. First, a fully convolutional network (FCN) was trained and assessed using accuracy and recall/precision metrics, and the early stopping technique was used to avoid overfitting. To assess the performance, the FCN was evaluated in terms of various metrics, including accuracy, precision, recall, F1-score, and receiver operating characteristic (ROC)-area under the curve (AUC). In addition, two-dimensional (2D) convolutional neural networks (CNNs) and long short-term memory (LSTM) models were used to model the database, and their performance was thoroughly measured using different metrics, graphs, and confusion matrices. Using LSTM variants, such as standard LSTM, bidirectional LSTM, stacked LSTM, and LSTM attention mechanisms, hybrid convolutional LSTM (ConvLSTM) models were trained and compared. The comparison was conducted based on the training and validation accuracy and loss, as well as the graphs resulting from the precision–recall curves. Apart from DL approaches, EEG signal analysis using time–frequency techniques, such as wavelet transform and short-time Fourier transform, has also been investigated. These methods assisted in the analysis of the time–frequency features of EEG signals in combination with DL models. This study demonstrates that the performance of wearable EEG devices can be augmented using a combination of DL and seizure signal processing techniques. The FCN achieved an accuracy of 92%, with a recall for seizures of 33%, an F1-score of 0.03, and strong ROC-AUC results. The 2D CNN achieved 96% accuracy, a seizure recall of 70%, an F1-score of 0.12, and an ROC-AUC score of 78%. The baseline LSTM struggled with effectiveness at 53% accuracy with a seizure recall of 18%. In contrast, the LSTM model, which incorporated synthetic minority oversampling technique (SMOTE) balancing, was able to reach improvements of up to 89% accuracy, with a precision of 91%, a recall of 86%, an F1-score of 0.89, and a strong ROC curve performance. Among the models, the LSTM with SMOTE was the best performer, with 89% accuracy, 91% precision, 86% recall, and an F1-score of 0.89. These results provide evidence that applying techniques for data balancing in combination with certain DL network architectures significantly improves the detection of seizures using wearable EEG devices worn on the body. We believe that real-time monitoring and high-performance systems are feasible using optimized DL frameworks. The analysis of the performance of different models allows for the understanding of the possibilities of optimizing the architectures of DL algorithms for the modern diagnosis of epilepsy in real time. The source code used to carry out the experiments is publicly available at CHB-MIT EEG Dataset Python Scripts (https://www.kaggle.com/code/adnankust/adnaneeg1).

People with disabilities suffer from inability to communicate with their surroundings, so Human-Computer Interaction (HCI) technologies are used to have a means of communication for people with disabilities with their surroundings. HCI is an emerging technology in the disciplines of Artificial Intelligence and Biomedical Engineering. To power an external device, HCI technology uses several basic signals such as ECG, EMG, and EEG. Electrooculography (EOG) is a technique for measuring the potential difference between the cornea and the retina located between the front and back of the human eye, and the main application of EOG is to determine the directions of different eye movements. This study aims to assess eye movement for communication by persons with disabilities using electrocardiogram (EOG) data. In this study, the Supporting Vector Machine (SVM) and Long- Short term memory (LSTM) classification techniques was used and two types of features (statistical and time domain features) were used. Classification accuracy was 90.7% and 93.9% when using SVM with statistical domain and time domain features, respectively ,whereas Classification accuracy was 90.1% when using LSTM .

Accurate rainfall forecasting is crucial in activities such as agricultural planning, water resource management, and flood preparedness, so failure to accurately forecast rainfall might undermine these activities. This study aimed to improve the accuracy of daily rainfall prediction in Hai district, Kilimanjaro region, Tanzania, a region characterised by a bimodal rainfall pattern, by comparing rainfall prediction accuracies of Traditional Machine Learning (TML) and Deep Learning (DL) models. The 25-year daily rainfall dataset for Hai district was first pre-processed by cleaning, normalisation, and splitting into training (first 60%), validation (next 20%), and test (last 20%) sets. TML models (K-Nearest Neighbours (KNN) through its variant KNN Regressor and Support Vector Machine (SVM) through its variant Support Vector Regressor (SVR)) were developed, trained, and their performances compared with developed and trained DL models (Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU)). Each model’s test performance was evaluated using the Root Mean Squared Error (RMSE) metric. The findings revealed that the DL models, LSTM and GRU, achieved test RMSE scores of 6.35 mm and 6.49 mm, respectively, while the TML models, KNN and SVM, achieved test RMSE scores of 6.61 mm and 10.42 mm, respectively. The study demonstrated that DL models are more effective for accurate rainfall forecasting than TML models in regions with complex rainfall patterns. The insights gained will guide the development of more reliable, location-specific weather prediction systems, ultimately supporting informed decision-making in agriculture, environmental management, and flood risk reduction in Tanzania and other regions facing similar climatic challenges

Tanzania heavily depends on United States Dollar (USD) foreign currency to import various goods and services into the country. Failure to correctly forecast exchange rates between USD and Tanzanian Shillings (TZS) may pose risks such as inability to import intended goods and services, possibility of losing money in stock exchange markets and other investment businesses in case of unexpected currency appreciation or depreciation as well as poor investment decisions in foreign exchange markets. To address this, this study has developed and comparatively evaluated performances of LSTM (Long Short-Term Memory) and BiLSTM (Bidirectional LSTM) deep learning models for forecasting daily USD to TZS exchange rates. The findings reveal that, BiLSTM model outperforms LSTM in forecasting daily USD to TZS exchange rates, achieving a MAPE (Mean Absolute Percentage Error) score of 0.363 on test set (unseen data) compared to a MAPE score of 1.471 achieved by LSTM model. This study recommends to the prospective Artificial Intelligence (AI) researchers and software developers to use BiLSTM instead of LSTM model to forecast (predict) USD to TZS exchange rates. Also, this study has developed USD to TZS exchange rates dataset which can be used by AI researchers, saving them time and costs involved with creating datasets from scratch. This study has also developed ready to use BiLSTM and LSTM models which can be used by Tanzanian business men and women involved in stock exchange markets, foreign exchange markets and other businesses, to predict daily USD to TZS exchange rates and make appropriate business and investment decisions.

Industrial robots are widely used in critical tasks such as assembly, welding, and material handling as core components of modern manufacturing systems. For the reliable operation of these systems, early and accurate detection of execution failures is crucial. In this study, a comprehensive comparison of machine learning and deep learning methods is conducted for the classification of robot execution failures using data acquired from force-torque sensors. Three different feature engineering approaches are proposed. The first is a Baseline approach that includes 90 raw time-series features. The second is the Domain-6 approach, which consists of 6 basic statistical features per sensor (36 in total). The third is the Domain-12 approach, which comprises 12 comprehensive statistical features per sensor (72 in total). The domain features include the mean, standard deviation, minimum, maximum, range, slope, median, skewness, kurtosis, RMS, energy, and IQR. In total, ten classification algorithms are evaluated, including eight machine learning methods and two deep learning models: Support Vector Machines (SVM), Random Forest (RF), k-Nearest Neighbors (KNN), Artificial Neural Network (ANN), Naive Bayes (NB), Decision Trees (DT), eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM-LGBM), as well as a One-Dimensional Convolutional Neural Network (CNN-1D) and Long Short-Term Memory (LSTM). For traditional machine learning algorithms, 5 × 5 nested cross-validation is used, whereas for deep learning models, 5-fold cross-validation with a 20% validation split is employed. To ensure statistical reliability, all experiments are repeated over 30 independent runs. The experimental results demonstrate that feature engineering has a decisive impact on classification performance. In addition, regardless of the feature set, the highest accuracy (93.85% ± 0.90) is achieved by the Naive Bayes classifier using the Baseline features. The Domain-12 feature set provides consistent improvements across many algorithms, with substantial performance gains. The results are reported using accuracy, precision, recall, and F1-score metrics and are supported by confusion matrices. Finally, permutation feature importance analysis indicates that the skewness features of the Fx and Fy sensors are the most critical variables for failure detection. Overall, these findings show that time-domain statistical features offer an effective approach for robot failure classification.

This paper devoted to the optimization of surface electromyography (sEMG) based hand gesture recognition classifier model. Twenty-two features in time domain, frequency domain and time-frequency domain were extracted from sEMG of 6 muscles to classify 13 hand gestures. Correlation analysis was used to screen the features. Random forest (RF) and principal component analysis (PCA) feature selection were used to further reduce the feature dimensions. Finally, mean absolute value (MAV), root mean square (RMS), slope sign change (SSC), mean frequency (MNF), Willison amplitude (WAMP) and autoregressive coefficients (AR1), totally 6 features were selected. LDA (Linear Discriminant Analysis), SVM (Support Vector Machine), KNN (k-Nearest Neighbor), DT (Decision Tree) and NN (Neural Network), in total of five classification methods were applied. The accuracy of the classifier had no significant reduction, but the dimension of the feature set was reduced from 132 to 36, and the training time was greatly reduced. The result indicated that the combination of correlation analysis and random forest feature selection can optimize the classifier model and greatly reduce the training time of the classifier.

Recently, with the advent of various Internet of Things (IoT) applications, a massive amount of network traffic is being generated. A network operator must provide different quality of service, according to the service provided by each application. Toward this end, many studies have investigated how to classify various types of application network traffic accurately. Especially, since many applications use temporary or dynamic IP or Port numbers in the IoT environment, only payload-based network traffic classification technology is more suitable than the classification using the packet header information as well as payload. Furthermore, to automatically respond to various applications, it is necessary to classify traffic using deep learning without the network operator intervention. In this study, we propose a traffic classification scheme using a deep learning model in software defined networks. We generate flow-based payload datasets through our own network traffic pre-processing, and train two deep learning models: 1) the multi-layer long short-term memory (LSTM) model and 2) the combination of convolutional neural network and single-layer LSTM models, to perform network traffic classification. We also execute a model tuning procedure to find the optimal hyper-parameters of the two deep learning models. Lastly, we analyze the network traffic classification performance on the basis of the F1-score for the two deep learning models, and show the superiority of the multi-layer LSTM model for network packet classification.

ObjectiveThe objective of this study was to investigate a more accurate and efficient technique for assessing spasticity in stroke patients via surface electromyography (sEMG).Methods45 hemiplegic individuals were recruited and spasticity was assessed via the modified Ashworth scale (MAS). Multichannel sEMG data were collected from 3 muscles: the long head of the biceps brachii (LB), the short head of the biceps brachii (SB), and the brachioradialis (BR). Both time-domain and frequency-domain features were extracted. A K-nearest neighbour (k-NN) classifier was used to develop a new feature vector consisting of multichannel sEMG features. Finally, a model using this new feature was constructed and evaluated for classification accuracy.ResultsData from 40 patients were analysed, revealing significant correlations between MAS scores and sEMG features. Specifically, MAS exhibited strong positive correlations with 3 time-domain features: root mean square (RMS), integral sEMG (iEMG), and envelope area (EA) (r > 0.7). In contrast, frequency-domain features were negatively correlated with the MAS score (r < −0.7). A single-channel model and a single-feature model were developed as baselines. A k-NN classifier using a novel feature vector – integrating single-channel and single-feature data – enabled automatic spasticity grading, surpassing the performance of the baseline models. The proposed multichannel sEMG feature fusion model achieved an average accuracy of 78.7%, significantly outperforming both the single-channel model (LB: 66.0%, SB: 64.3%, BR: 70.4%) and the single-feature model (RMS 70.8%, iEMG 71.4%, and EA 63.4%).ConclusionsCompared with single-channel and single-feature models, the k-NN model, which uses multichannel sEMG features, has superior accuracy in spasticity assessments and is a reliable tool for objective evaluation. This approach holds promise for enhancing rehabilitation strategies by enabling precise and data-driven efficacy assessments, ultimately improving patient outcomes.