Convolutional Neural Network Long Short-term Memory Research Articles

Traffic prediction via clustering and deep transfer learning with limited data

AbstractThis paper proposes a method based on the clustering algorithm, deep learning, and transfer learning (TL) for short‐term traffic prediction with limited data. To address the challenges posed by limited data and the complex and diverse traffic patterns observed in traffic networks, we propose a profile model based on few‐shot learning to extract each detector's unique profiles. These profiles are then used to cluster detectors with similar patterns into distinct clusters, facilitating effective learning with limited data. A Convolutional Neural Network ‐ Long Short‐Term Memory (CNN‐LSTM)‐based predictive model is proposed to learn and predict traffic volumes for each detector within a cluster. The proposed method demonstrates resilience to detector failures and has been validated using the Performance Measurement System dataset. In scenarios with less than 2 months of training data and 10% failed detectors, the mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) for station‐level traffic volume prediction increase from 12.7 vehs/5 min, 20.9 vehs/5 min, and 10.5% to 13.9 vehs/5 min, 24.2 vehs/5 min, and 11.7%, respectively. For lane‐level traffic volume prediction, the average MAE, RMSE, and MAPE increase from 4.7 vehs/5 min, 7.7 vehs/5 min, and 15% to 5.6 vehs/5 min, 9.6 vehs/5 min, and 16.8%. Furthermore, the proposed method extends its applicability to traffic speed and occupancy prediction tasks. TL is integrated to improve speed/occupancy prediction accuracy by leveraging knowledge obtained from traffic volume, considering the correlation between traffic flow, speed, and occupancy. When less than 1 month of traffic speed/occupancy data is available for learning, the proposed method achieves an MAE, RMSE, and MAPE of 0.7 km/h, 1.3 km/h, and 1.3% for station‐level traffic speed prediction and 0.5%, 1.1%, and 11% for station‐level traffic occupancy.

Enhanced Forecasting Accuracy of a Grid-Connected Photovoltaic Power Plant: A Novel Approach Using Hybrid Variational Mode Decomposition and a CNN-LSTM Model

Renewable energies have become pivotal in the global energy landscape. Their adoption is crucial for phasing out fossil fuels and promoting environmentally friendly energy solutions. In recent years, the energy management system (EMS) concept has emerged to manage the power grid. EMS optimizes electric grid operations through advanced metering, automation, and communication technologies. A critical component of EMS is power forecasting, which facilitates precise energy grid scheduling. This research paper introduces a deep learning hybrid model employing convolutional neural network–long short-term memory (CNN-LSTM) for short-term photovoltaic (PV) solar energy forecasting. The proposed method integrates the variational mode decomposition (VMD) algorithm with the CNN-LSTM model to predict PV power output from a solar farm in Boussada, Algeria, spanning 1 January 2019, to 31 December 2020. The performance of the developed model is benchmarked against other deep learning models across various time horizons (15, 30, and 60 min): variational mode decomposition–convolutional neural network (VMD-CNN), variational mode decomposition–long short-term memory (VMD-LSTM), and convolutional neural network–long short-term memory (CNN-LSTM), which provide a comprehensive evaluation. Our findings demonstrate that the developed model outperforms other methods, offering promising results in solar power forecasting. This research contributes to the primary goal of enhancing EMS by providing accurate solar energy forecasts.

Human motion activity recognition and pattern analysis using compressed deep neural networks

ABSTRACT This work presents an on-device machine learning model with the ability to identify different mobility gestures called human activity recognition (HAR), which includes running, walking, squatting, jumping, and others. The data is collected through an Arduino Nano 33 BLE Sense board with a sampling rate of 119 Hz, which is embedded with an Inertial Measurement Unit (IMU) sensor. The same board is used as a microcontroller to identify human gestures by developing an end-to-end edge computing application. A deep neural network model is trained and then compressed for deployment on the board to create a self-contained, embedded device capable of identifying the type of gesture performed. Three deep learning models, namely Multi-Layer Perceptron (MLP), Convolutional Neural Network – Long Short Term Memory (CNN-LSTM) & CNN-Gated Recurrent Unit (CNN-GRU), are evaluated in the identification of the mobility gestures. The observed accuracy of the models is 96%, 97.1% and 97.8%, respectively, MLP, CNN-LSTM & CNN-GRU across the different gesture categories. The study shows the utility of embedded devices with deep neural network-based models, which can provide low cost, minimal power usage, and meet data privacy requirements in HAR.

IoT Threat Mitigation: Leveraging Deep Learning for Intrusion Detection

The growth of smart gadgets connected via the Internet of Things (IoT) in today’s modern technology landscape has substantially improved our everyday lives. However, this convenience is juxtaposed with a concomitant surge in cyber threats capable of compromising the integrity of these interconnected systems. Conventional intrusion detection systems (IDS) prove inadequate for IoT due to the unique challenges they present. We propose and evaluate an intrusion detection system (IDS) based on a hybrid Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) model in this paper. The model is designed to capture both temporal and spatial patterns in network data, offering a robust solution for detecting malicious activities within IoT environments. The CNN-LSTM model displayed excellent accuracy, reaching 98% in both multi-class and binary classifications when trained on the UNSW-NB15 dataset. Furthermore, we explore the real-world applicability of the model through testing on Raspberry Pi, showcasing its effectiveness in IoT scenarios. The system is augmented with alert mechanisms, promptly notifying relevant parties upon intrusion detection. Our findings highlight the CNN-LSTM model's efficacy in strengthening IoT network security.

CNN–LSTM–AM: A power prediction model for offshore wind turbines

This study introduces a power forecasting model, the convolutional neural network (CNN)–long short-term memory (LSTM)–attention mechanism (AM) algorithm (CNN–LSTM–AM), designed to predict the power of offshore wind turbines based on data collected by a SCADA system. The model employs a timestep parameterisation approach for offshore wind turbine prediction, facilitating automatic partitioning of the training dataset and simplifying the training process. A CNN–LSTM–AM network was presented to predict the power of offshore wind turbines using signals from multiple sensors. A variable–control comparison was conducted to complete the sensitivity analysis of the sensors, which determined the most suitable sensor group for power prediction. The model achieved a maximum improvement of 13.77% in power prediction compared to existing deep learning algorithms. The results indicate that the hub and rear-end temperatures of the high-speed shaft of the gearboxes are crucial for offshore wind power prediction. Overall, the findings of this study contribute to the operation and maintenance of offshore wind turbines and the management of offshore wind farms.

Comparison of Water Quality Prediction for Red Tilapia Aquaculture in an Outdoor Recirculation System Using Deep Learning and a Hybrid Model

In modern aquaculture, the focus is on optimizing production and minimizing environmental impact through the use of recirculating water systems, particularly in outdoor setups. In such systems, maintaining water quality is crucial for sustaining a healthy environment for aquatic life, and challenges arise from instrumentation limitations and delays in laboratory measurements that can impact aquatic animal production. This study aimed to predict key water quality parameters in an outdoor recirculation aquaculture system (RAS) for red tilapia aquaculture, including dissolved oxygen (DO), pH, total ammonia nitrogen (TAN), nitrite nitrogen (NO2–N), and alkalinity (ALK). Initially, a random forest (RF) model was employed to identify significant factors for predicting each parameter, selecting the top three features from routinely measured parameters on the farm: DO, pH, water temperature (Temp), TAN, NO2–N, and transparency (Trans). This approach aimed to streamline the analysis by reducing variables and computation time. The selected parameters were then used for prediction, comparing the performance of convolutional neural network (CNN), long short-term memory (LSTM), and CNN–LSTM models across different epochs (1000, 3000, and 5000). The results indicated that the CNN–LSTM model at 5000 epochs was effective in predicting DO, TAN, NO2–N, and ALK, with high R2 values (0.815, 0.826, 0.831, and 0.780, respectively). However, pH prediction showed lower efficiency with an R2 value of 0.377.

Multimodel ensemble estimation of Landsat-like global terrestrial latent heat flux using a generalized deep CNN-LSTM integration algorithm

Accurate estimates of high-spatial-resolution global terrestrial latent heat flux (LE) from Landsat data are crucial for many basic and applied research. Yet current Landsat-derived LE products were developed using single algorithm with large uncertainties and discrepancies. Here we proposed a convolutional neural network-long short-term memory (CNN-LSTM)-based integrated LE (CNN-LSTM-ILE) framework that integrates five Landsat-derived physical LE algorithms, topography-related variables (elevation, slope and aspect) and eddy covariance (EC) observations to estimate 30-m global terrestrial LE. CNN-LSTM-ILE not only conserves good performance of LE estimation from pure deep learning (DL) algorithm, but partially inherits physical mechanism of the individual physical algorithms for improving the generalization of the integration algorithms for extreme cases. CNN-LSTM is an algorithm that combines two deep learning structures (CNN and LSTM) to effectively utilize the spatial and temporal information contained in the forcing inputs. The data were collected from 190 globally distributed EC observations from 2000 to 2015 and were provided by FLUXNET. The cross-validation results indicated that the CNN-LSTM integration algorithm improved the LE estimates by reducing the root mean square error (RMSE) of 5–8 W/m2 and increasing Kling-Gupta efficiency (KGE) of 0.05–0.16 when compared with the individual LE algorithms and the results of three other machine learning integration algorithms (multiple linear regression, MLR; random forest, RF; and deep neural networks, DNN). The CNN-LSTM integration algorithm had highest KGE (0.81) and R2 (0.66) compared to ground-measured and was applied to generate the Landsat-like regional and global terrestrial LE. An innovation of our strategy is that the CNN-LSTM-ILE model integrates pixel proximity effects and daily LE variations to enhance the accuracy of 16-day LE estimations. This approach can produce a more reliable Landsat-like global terrestrial LE product to improve the representativeness of heterogeneous regions for monitoring hydrological variables.

W-WaveNet: A multi-site water quality prediction model incorporating adaptive graph convolution and CNN-LSTM.

Water quality prediction is of great significance in pollution control, prevention, and management. Deep learning models have been applied to water quality prediction in many recent studies. However, most existing deep learning models for water quality prediction are used for single-site data, only considering the time dependency of water quality data and ignoring the spatial correlation among multi-sites. This research defines and analyzes the non-aligned spatial correlations that exist in multi-site water quality data. Then deploy spatial-temporal graph convolution to process water quality data, which takes into account both the temporal and spatial correlation of multi-site water quality data. A multi-site water pollution prediction method called W-WaveNet is proposed that integrates adaptive graph convolution and Convolutional Neural Network, Long Short-Term Memory (CNN-LSTM). It integrates temporal and spatial models by interleaved stacking. Theoretical analysis shows that the method can deal with non-aligned spatial correlations in different time spans, which is suitable for water quality data processing. The model validates water quality data generated on two real river sections that have multiple sites. The experimental results were compared with the results of Support Vector Regression, CNN-LSTM, and Spatial-Temporal Graph Convolutional Networks (STGCN). It shows that when W-WaveNet predicts water quality over two river sections, the average Mean Absolute Error is 0.264, which is 45.2% lower than the commonly used CNN-LSTM model and 23.8% lower than the STGCN. The comparison experiments also demonstrate that W-WaveNet has a more stable performance in predicting longer sequences.

Spatio-Temporal Deep Learning-Based Forecasting of Surface Solar Irradiance: Leveraging Satellite Data and Feature Selection

Spatio-Temporal Deep Learning-Based Forecasting of Surface Solar Irradiance: Leveraging Satellite Data and Feature Selection

Air gap flux density measurement of PMSLM based on TMR sensing external stray magnetic field and CNN-LSTM

Air gap flux density measurement of PMSLM based on TMR sensing external stray magnetic field and CNN-LSTM

Deep learning based biometric authentication using electrocardiogram and iris

Authentication systems play an important role in wide range of applications. The traditional token certificate and password-based authentication systems are now replaced by biometric authentication systems. Generally, these authentication systems are based on the data obtained from face, iris, electrocardiogram (ECG), fingerprint and palm print. But these types of models are unimodal authentication, which suffer from accuracy and reliability issues. In this regard, multimodal biometric authentication systems have gained huge attention to develop the robust authentication systems. Moreover, the current development in deep learning schemes have proliferated to develop more robust architecture to overcome the issues of tradition machine learning based authentication systems. In this work, we have adopted ECG and iris data and trained the obtained features with the help of hybrid convolutional neural network- long short-term memory (CNN-LSTM) model. In ECG, R peak detection is considered as an important aspect for feature extraction and morphological features are extracted. Similarly, gabor-wavelet, gray level co-occurrence matrix (GLCM), gray level difference matrix (GLDM) and principal component analysis (PCA) based feature extraction methods are applied on iris data. The final feature vector is obtained from MIT-BIH and IIT Delhi Iris dataset which is trained and tested by using CNN-LSTM. The experimental analysis shows that the proposed approach achieves average accuracy, precision, and F1-core as 0.985, 0.962 and 0.975, respectively.

Implementing deep learning-based named entity recognition for obtaining narcotics abuse data in Indonesia

<span lang="EN-US">The availability of drug abuse data from the official website of the National Narcotics Board of Indonesia is not up-to-date. Besides, the drug reports from Indonesian National Narcotics Board are only published once a year. This study aims to utilize online news sites as a data source for collecting information about drug abuse in Indonesia. In addition, this study also builds a named entity recognition (NER) model to extract information from news texts. The primary NER model in this study uses the convolutional neural network-long short-term memory (CNNs-LSTM) architecture because it can produce a good performance and only requires a relatively short computation time. Meanwhile, the baseline NER model uses the bidirectional long short-term memory-conditional random field (Bi-LSTMs-CRF) architecture because it is easy to implement using the Flair framework. The primary model that has been built results in a performance (F1 score) of 82.54%. Meanwhile, the baseline model only results in a performance (F1 score) of 69.67%. Then, the raw data extracted by NER is processed to produce the number of drug suspects in Indonesia from 2018-2020. However, the data that has been produced is not as complete as similar data sourced from Indonesian National Narcotics Board publications.</span>

Deep learning modeling of human activity affected wildfire risk by incorporating structural features: A case study in eastern China

Wildfire risk prediction is a critical component of disaster prevention and mitigation, often closely associated with local human activities in most regions. Recent studies demonstrate that employing joint modeling techniques using diverse datasets alongside Convolutional Neural Networks-Long Short-Term Memory Networks (ConvLSTM) produces favorable predictive results. However, previous research inadequately explored the different impact of factors across different categories and spatial orientations, and neglected the impact of fuels and human activities inside the samples. This study focuses on the six eastern provinces of China, utilizing a multi-source dataset comprising satellite-monitored wildfire products from 2012 to 2022, along with various factors indicating terrestrial and human activities, simulated meteorological elements and high-resolution vegetation imagery. By introducing channel and spatial attention mechanisms and visual transformer mode, this research optimizes the ConvLSTM wildfire prediction model. Results indicate a noteworthy enhancement, elevating accuracy, Kappa coefficient, and AUC of ROC curves from 91.15%, 80.87%, and 97.01% to 92.79%, 84.48%, and 97.90%, respectively. Consequently, it reinforces the accuracy by increase of the structural features within samples and quantifying the differences in the importance of different factors, which is also validated by prediction application of the samples in the entire year of 2023. Sensitivity analysis reveals that the current model is still highly dependent on the meteorological factors. Notably, the impact of structural features significantly surpasses the influence of terrain and terrestrial ecology elements, which should be considered in further models. Thus, this study has developed a methodology integrating multiple attention mechanisms and sample structural features, which could furnish high-precision daily kilometer-level wildfire risk prediction products. This method could improve the efficiency of prevention and control by improving the accuracy and narrowing the high-risk areas.

Comparison of simulation and predictive efficacy for hemorrhagic fever with renal syndrome incidence in mainland China based on five time series models.

Hemorrhagic fever with renal syndrome (HFRS) is a zoonotic infectious disease commonly found in Asia and Europe, characterized by fever, hemorrhage, shock, and renal failure. China is the most severely affected region, necessitating an analysis of the temporal incidence patterns in the country. We employed Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), Nonlinear AutoRegressive with eXogenous inputs (NARX), and a hybrid CNN-LSTM model to model and forecast time series data spanning from January 2009 to November 2023 in the mainland China. By comparing the simulated performance of these models on training and testing sets, we determined the most suitable model. Overall, the CNN-LSTM model demonstrated optimal fitting performance (with Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Mean Absolute Error (MAE) of 93.77/270.66, 7.59%/38.96%, and 64.37/189.73 for the training and testing sets, respectively, lower than those of individual CNN or LSTM models). The hybrid CNN-LSTM model seamlessly integrates CNN's data feature extraction and LSTM's recurrent prediction capabilities, rendering it theoretically applicable for simulating diverse distributed time series data. We recommend that the CNN-LSTM model be considered as a valuable time series analysis tool for disease prediction by policy-makers.

ECG-based cardiac arrhythmias detection through ensemble learning and fusion of deep spatial–temporal and long-range dependency features

Cardiac arrhythmia is one of the prime reasons for death globally. Early diagnosis of heart arrhythmia is crucial to provide timely medical treatment. Heart arrhythmias are diagnosed by analyzing the electrocardiogram (ECG) of patients. Manual analysis of ECG is time-consuming and challenging. Hence, effective automated detection of heart arrhythmias is important to produce reliable results. Different deep-learning techniques to detect heart arrhythmias such as Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), Transformer, and Hybrid CNN–LSTM were proposed. However, these techniques, when used individually, are not sufficient to effectively learn multiple features from the ECG signal. The fusion of CNN and LSTM overcomes the limitations of CNN in the existing studies as CNN–LSTM hybrids can extract spatiotemporal features. However, LSTMs suffer from long-range dependency issues due to which certain features may be ignored. Hence, to compensate for the drawbacks of the existing models, this paper proposes a more comprehensive feature fusion technique by merging CNN, LSTM, and Transformer models. The fusion of these models facilitates learning spatial, temporal, and long-range dependency features, hence, helping to capture different attributes of the ECG signal. These features are subsequently passed to a majority voting classifier equipped with three traditional base learners. The traditional learners are enriched with deep features instead of handcrafted features. Experiments are performed on the MIT-BIH arrhythmias database and the model performance is compared with that of the state-of-art models. Results reveal that the proposed model performs better than the existing models yielding an accuracy of 99.56%.

Hybrid deep learning models for time series forecasting of solar power

AbstractForecasting solar power production accurately is critical for effectively planning and managing renewable energy systems. This paper introduces and investigates novel hybrid deep learning models for solar power forecasting using time series data. The research analyzes the efficacy of various models for capturing the complex patterns present in solar power data. In this study, all of the possible combinations of convolutional neural network (CNN), long short-term memory (LSTM), and transformer (TF) models are experimented. These hybrid models also compared with the single CNN, LSTM and TF models with respect to different kinds of optimizers. Three different evaluation metrics are also employed for performance analysis. Results show that the CNN–LSTM–TF hybrid model outperforms the other models, with a mean absolute error (MAE) of 0.551% when using the Nadam optimizer. However, the TF–LSTM model has relatively low performance, with an MAE of 16.17%, highlighting the difficulties in making reliable predictions of solar power. This result provides valuable insights for optimizing and planning renewable energy systems, highlighting the significance of selecting appropriate models and optimizers for accurate solar power forecasting. This is the first time such a comprehensive work presented that also involves transformer networks in hybrid models for solar power forecasting.

Estimation of Muscle Forces of Lower Limbs Based on CNN-LSTM Neural Network and Wearable Sensor System.

Estimation of vivo muscle forces during human motion is important for understanding human motion control mechanisms and joint mechanics. This paper combined the advantages of the convolutional neural network (CNN) and long-short-term memory (LSTM) and proposed a novel muscle force estimation method based on CNN-LSTM. A wearable sensor system was also developed to collect the angles and angular velocities of the hip, knee, and ankle joints in the sagittal plane during walking, and the collected kinematic data were used as the input for the neural network model. In this paper, the muscle forces calculated using OpenSim based on the Static Optimization (SO) method were used as the standard value to train the neural network model. Four lower limb muscles of the left leg, including gluteus maximus (GM), rectus femoris (RF), gastrocnemius (GAST), and soleus (SOL), were selected as the studying objects in this paper. The experiment results showed that compared to the standard CNN and the standard LSTM, the CNN-LSTM performed better in muscle forces estimation under slow (1.2 m/s), medium (1.5 m/s), and fast walking speeds (1.8 m/s). The average correlation coefficients between true and estimated values of four muscle forces under slow, medium, and fast walking speeds were 0.9801, 0.9829, and 0.9809, respectively. The average correlation coefficients had smaller fluctuations under different walking speeds, which indicated that the model had good robustness. The external testing experiment showed that the CNN-LSTM also had good generalization. The model performed well when the estimated object was not included in the training sample. This article proposed a convenient method for estimating muscle forces, which could provide theoretical assistance for the quantitative analysis of human motion and muscle injury. The method has established the relationship between joint kinematic signals and muscle forces during walking based on a neural network model; compared to the SO method to calculate muscle forces in OpenSim, it is more convenient and efficient in clinical analysis or engineering applications.

Open space radar specific emitter identification using MSAK‐CNN‐LSTM network

AbstractTo enhance the capability of identifying unknown emitters in open spaces, an open‐multiscale attention kernel (MSAK)‐convolutional neural network‐long short‐term memory (CNN‐LSTM) structure is proposed. To this end, first, a MSAK module and CNN‐LSTM structure are introduced, and then, the depth and complexity of the feature extraction network are improved to enhance its representation capability. To classify unknown emitters accurately, the MSAK‐CNN‐LSTM model is improved to obtain an open‐MSAK‐CNN‐LSTM model with open‐set recognition capability. Additionally, the two preprocessing procedures are summarised, and their strengths and weaknesses are compared. Experimental results show that the proposed open‐MSAK‐CNN‐LSTM model achieves satisfactory accuracy in identifying unknown emitters in open space. In addition, it has significant advantages in low signal‐to‐noise ratio (SNR) scenarios.

Comparative analysis of machine learning techniques for cryptocurrency price prediction

The emergence of cryptocurrencies has revolutionized the concept of digital currencies and attracted significant attention from financial markets. Predicting the price dynamics of cryptocurrencies is crucial but challenging due to their highly volatile and non-linear nature. This study compares the performance of various models in predicting cryptocurrency prices using three datasets: Bitcoin (BTC), Litecoin (LTC), and Ethereum (ETH). The models analyzed include Moving Average (MA), Logistic Regression (LR), Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), and Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM). The objective is to uncover underlying patterns in cryptocurrency price movements and identify the most accurate and reliable approach for predicting future prices. Through the analysis, it could be observed that MA, LR, and ARIMA models struggle to capture the actual trend accurately. In contrast, LSTM and CNN-LSTM models demonstrate strong fit to the actual price trend, with CNN-LSTM exhibiting a higher level of granularity in its predictions. Results suggest that deep learning architectures, particularly CNN-LSTM, show promise in capturing the complex dynamics of cryptocurrency prices. These findings contribute to the development of improved methodologies for cryptocurrency price prediction.

Comparative study of univariate and multivariate strategy for short-term forecasting of heat demand density: Exploring single and hybrid deep learning models

Accurate short-term forecasting of heating energy demand is needed for achieving optimal building energy management, cost savings, environmental sustainability, and responsible energy consumption. Furthermore, short-term heating energy prediction contributes to zero-energy building performance in cold climates. Given the critical importance of short-term forecasting in heating energy management, this study evaluated six prevalent deep-learning algorithms to predict energy load, including single and hybrid models. The overall best-performing predictors were hybrid models using Convolutional Neural Networks, regardless of whether they were multivariate or univariate. Nevertheless, while the multivariate models performed better in the first hour, the univariate models often were more accurate in the final 24 h. Thus, the best-performing predictor of the first timestep was a multivariate hybrid Convolutional Neural Network–Recurrent Neural Network model with a coefficient of determination (R²) of 0.98 and the lowest mean absolute error. Yet, the best-performing predictor of the final timestep was the univariate hybrid model Convolutional Neural Network–Long Short-Term Memory with an R² of 0.80. Also, the prediction accuracy of the best-performing multivariate hybrid models reduced faster per hour compared to the univariate models. These findings suggest that multivariate models may be better suited for early timestep predictions, while univariate models may be better suited for later time steps. Hence, combining the models can enhance accuracy at various timesteps for achieving high fidelity in forecasting and offering a comprehensive tool for energy management.