Long Short-term Memory Deep Learning Research Articles

LSTM Deep Learning Techniques for Wind Power Generation Forecasting

Wind power generation forecasting is crucial for the optimal integration of renewable energy sources into power systems. Traditional forecasting methods often struggle to accurately predict wind energy production due to the complex and nonlinear relationships between wind speed, weather parameters, and power output. In recent years, deep learning techniques have emerged as promising alternatives for wind power forecasting. This conference paper provides a comprehensive review of deep learning techniques, with a specific focus on Long Short-Term Memory (LSTM) networks, for short-term wind power generation forecasting. Leveraging insights from recent research and empirical evaluations, this paper explores the effectiveness of LSTM networks in capturing temporal dependencies in wind data and improving prediction accuracy. The review highlights the potential of LSTM-based models to enhance the integration of wind energy into power systems and provides guidance for future research in this area.

Cortical signals analysis to recognize intralimb mobility using modified RNN and various EEG quantities

Electroencephalogram (EEG) signals are critical in interpreting sensorimotor activities for predicting body movements. However, their efficacy in identifying intralimb movements, such as the dorsiflexion and plantar flexion of the foot, remains suboptimal. This study aims to explore whether various EEG signal quantities can effectively recognize intralimb movements to facilitate the development of Brain-Computer Interface (BCI) devices for foot rehabilitation. This research involved twenty-two healthy, right-handed participants. EEG data were collected using 21 electrodes positioned over the motor cortex, while two electromyography (EMG) electrodes recorded the onset of ankle joint movements. The study focused on analyzing slow cortical potential (SCP) and sensorimotor rhythms (SMR) in alpha and beta bands from the EEG. Five key features—fourth-order Autoregressive feature, variance, waveform length, standard deviation, and permutation entropy—were extracted. A modified Recurrent Neural Network (RNN) including Long Short-term Memory (LSTM) and Gated Recurrent Unit (GRU) algorithms was developed for movement recognition. These were compared against conventional machine learning algorithms, including nonlinear Support Vector Machine (SVM) and k Nearest Neighbourhood (kNN) classifiers. The performance of the proposed models was assessed using two data schemes: within-subject and across-subjects. The findings demonstrated that the GRU and LSTM models significantly outperformed traditional machine learning algorithms in recognizing different EEG signal quantities for intralimb movement. The study indicates that deep learning models, particularly GRU and LSTM, hold superior potential over standard machine learning techniques in identifying intralimb movements using EEG signals. Where the accuracies of LSTM for within and across subjects were 98.87 ± 1.80 % and 87.38 ± 0.86 % respectively. Whereas the accuracy of GRU within and across subjects were 99.18 ± 1.28 % and 86.44 ± 0.69 % respectively. This advancement could significantly benefit the development of BCI devices aimed at foot rehabilitation, suggesting a new avenue for enhancing physical therapy outcomes.

Deep Learning Models for Bitcoin Prediction Using Hybrid Approaches with Gradient-Specific Optimization

Since cryptocurrencies are among the most extensively traded financial instruments globally, predicting their price has become a crucial topic for investors. Our dataset, which includes fluctuations in Bitcoin’s hourly prices from 15 May 2018 to 19 January 2024, was gathered from Crypto Data Download. It is made up of over 50,000 hourly data points that provide a detailed view of the price behavior of Bitcoin over a five-year period. In this study, we used potent algorithms, including gradient descent, attention mechanisms, long short-term memory (LSTM), and artificial neural networks (ANNs). Furthermore, to estimate the price of Bitcoin, we first merged two deep learning algorithms, LSTM and attention mechanisms, and then combined LSTM-Attention with gradient-specific optimization to increase our model’s performance. Then we integrated ANN-LSTM and included gradient-specific optimization for the same reason. Our results show that the hybrid model with gradient-specific optimization can be used to anticipate Bitcoin values with better accuracy. Indeed, the hybrid model combines the best features of both approaches, and gradient-specific optimization improves predictive performance through frequent analysis of pricing data changes.

A maximal overlap discrete wavelet packet transform coupled with an LSTM deep learning model for improving multilevel groundwater level forecasts

Developing precise groundwater level (GWL) forecast models is essential for the optimal usage of limited groundwater resources and sustainable planning and management of water resources. In this study, an improved forecasting accuracy for up to 3 weeks ahead of GWLs in Bangladesh was achieved by employing a coupled Long Short Term Memory (LSTM) network-based deep learning algorithm and Maximal Overlap Discrete Wavelet Packet Transform (MODWPT) data preprocessing. The coupled LSTM-MODWPT model’s performance was compared with that of the LSTM model. For both standalone LSTM and LSTM-MODWPT models, the Random Forest feature selection approach was employed to select the ideal inputs from the candidate GWL lags. In the LSTM-MODWPT model, input GWL time series were decomposed using MODWPT. The ‘Fejér-Korovkin’ mother wavelet with a filter length of 18 was used to obtain a collection of scaling coefficients and wavelets for every single input time series. Model performance was assessed using five performance indices: Root Mean Squared Error; Scatter Index; Maximum Absolute Error; Median Absolute Deviation; and an a-20 index. The LSTM-MODWPT model outperformed standalone LSTM models for all time horizons in GWL forecasting. The percentage improvements in the forecasting accuracies were 36.28%, 32.97%, and 30.77%, respectively, for 1-, 2-, and 3-weeks ahead forecasts at the observation well GT3330001. Accordingly, the coupled LSTM-MODWPT model could potentially be used to enhance multiscale GWL forecasts. This research demonstrates that the coupled LSTM-MODWPT model could generate more precise GWL forecasts at the Bangladesh study site, with potential applications in other geographic locations globally.

LSTM deep learning long-term traffic volume prediction model based on Markov state description

ABSTRACT When applying long short-term memory (LSTM) neural network model to traffic prediction, there are limitations in exploiting spatial-temporal traffic state features. The interpretability of models has not received enough attention. This study suggests an LSTM traffic flow prediction model that can anticipate traffic volume 24 h in advance. The model makes use of the traffic flow state information obtained from the fuzzy C-means clustering method by clustering the multi-day historical traffic flow data. Markov chain is used to capture the label feature of traffic flow using transition probability matrix information. To show the efficacy of the suggested technique, experiments were conducted using real traffic volume data from a city in China. The simulation results demonstrate that the proposed model can attain greater prediction accuracy, and the network training time may be significantly reduced.

Comparative Analysis of Machine Learning and Deep Learning Models for Bitcoin Price Prediction

This research endeavors to forecast Bitcoin prices by employing a suite of machine learning and deep learning models. Five distinct models were employed: Random Forest, Linear Regression, Support Vector Machine (SVM), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU), each evaluated based on their R-squared scores. Notably, the models showcased diverse performances, with the ensemble learning approach of Random Forest exhibiting near-perfect accuracy, closely followed by GRU and SVM. The deep learning architectures, LSTM and GRU, demonstrated remarkable predictive capabilities, showcasing their adeptness in capturing intricate temporal patterns within the cryptocurrency price data. This study sheds light on the comparative performance of these models, emphasizing their strengths and limitations in predicting Bitcoin prices.

A hybrid deep learning technique for spoofing website URL detection in real-time applications

Website Uniform Resource Locator (URL) spoofing remains one of the ways of perpetrating phishing attacks in the twenty-first century. Hackers continue to employ URL spoofing to deceive naïve and unsuspecting consumers into releasing important personal details in malicious websites. Blacklists and rule-based filters that were once effective at reducing the risks and sophistication of phishing are no longer effective as there are over 1.5 million new phishing websites created monthly. Therefore, research aimed at unveiling new techniques for detecting phishing websites has sparked a lot of interest in both academics and business with machine and deep learning techniques being at the forefront. Among the deep learning techniques that have been employed, Convolutional Neural Network (CNN) remains one of the most widely used with high performance in feature learning. However, CNN has a problem of memorizing contextual relationships in URL text, which makes it challenging to efficiently detect sophisticated malicious URLs in real-time applications. On the contrary, Long Short-Term Memory (LSTM) deep learning model has been successfully employed in complex real-time problems because of its ability to store inputs for a long period of time. This study experiments with the use of hybrid CNN and LSTM deep learning models for spoofing website URL detection in order to exploit the combined strengths of the two approaches for a more sophisticated spoofing URL detection. Two publicly available datasets (UCL spoofing Website and PhishTank Datasets) were used to evaluate the performance of the proposed hybrid model against other models in the literature. The hybrid CNN-LSTM model achieved accuracies of 98.9% and 96.8%, respectively, when evaluated using the UCL and PhishTank datasets. On the other hand, the standalone CNN and LSTM achieved accuracies of 90.4% and 94.6% on the UCL dataset, while their accuracies on the PhishTank dataset were 89.3% and 92.6%, respectively. The results show that the hybrid CNN-LSTM algorithm largely outperformed the standalone CNN and LSTM models, which demonstrates a much better performance. Therefore, the hybrid deep learning technique is recommended for detecting spoofing website URL thereby reducing losses attributed to such attacks.

Drought index downscaling using AI-based ensemble technique and satellite data

This study introduces and validates an artificial intelligence (AI)–based downscaling method for Standardized Precipitation Indices (SPI) in the northwest of Iran, utilizing PERSSIAN-CDR data and MODIS-derived drought-dependent variables. The correlation between SPI and two drought-dependent variables at a spatial resolution of 0.25° from 2000 to 2015 served as the basis for predicting SPI values at a finer spatial resolution of 0.05° for the period spanning 2016 to 2021. Shallow AI models (Support Vector Regression, Adaptive Neural Fuzzy Inference System, Feedforward Neural Network) and the Long Short-Term Memory (LSTM) deep learning method are employed for downscaling, followed by an ensemble post-processing technique for shallow AI models. Validation against rain gauge data indicates that all methods improve SPI simulation compared to PERSIANN-CDR products. The ensemble technique excels by 20% and 25% in the training and test phases, respectively, achieving the mean Determination Coefficient (DC) score of 0.67 in the validation phase. Results suggest that the deep learning LSTM method is less suitable for limited observed data compared to ensemble techniques. Additionally, the proposed methodology successfully detects approximately 80% of drought conditions. Notably, SPI-6 outperforms other temporal scales. This study advances the understanding of AI-driven downscaling for SPI, emphasizing the efficacy of ensemble approaches and providing valuable insights for regions with limited observational data.

A GMEE-WFED System: Optimizing Wind Turbine Distribution for Enhanced Renewable Energy Generation in the Future

This paper presents the Generation Max Electrical Energy from Wind Friendly Environment Database (GMEE-WFED) system, a groundbreaking innovation aimed at enhancement the performance and energy output of wind power generation stations. The GMEE-WFED system has been meticulously designed to provide precise wind power forecasting within distributed turbine systems, facilitating the seamless integration of renewable energy into the grid. This forecasting is enabled by the utilization of the Spatial Dynamic Wind Power Forecasting (SDWPF) dataset, which takes into account the spatial distribution and dynamic characteristics of wind turbines. The GMEE-WFED system comprises five layers, each offering unique advantages. The first layer, referred to as the "Best Distribution of Turbines Based on DOA (BDT-DOA)," is designed to achieve the following objectives: (a) increase power generation, (b) determine the optimal coordinates (x, y) for each turbine, and (c) distribute turbines based on the best locations. The second layer, named the "Effect Features Layer (EF)," focuses on: (a) identifying the impact of features on wind power generation, (b) streamlining implementation time, and (c) reducing computational demands. The third layer, denoted as the "Average and Shifting up Target Layer (AEH-SUV)," serves the purposes of: (a) enhancing accuracy by calculating feature averages, and (b) predicting future active power through target shifting at different intervals (ranging from 1 to 6 h). Meanwhile, the fourth layer is associated with the development of a prediction model based on a deep learning technique known as "Deep Learning-Long Short-Term Memory (DL-LSTM) Layer," which is utilized for: (a) forecasting future energy production, (b) evaluating model accuracy at varying intervals, and (c) assessing overall model effectiveness. The final layer is also dedicated to constructing a prediction model, but it leverages a different deep learning technique called "Deep Learning Gate Recurrent Unit (DL-GRU)." These models contribute to accurate wind power predictions at various intervals and ensure the overall effectiveness of the system. Experimental results have shown that DL-GRU outperforms DL-LSTM in all shifting cases, underscoring the system's effectiveness in predicting future wind power generation and forecasting accuracy. As a result, the GMEE-WFED system is a pioneering approach that enhances wind DC-power generation forecasting. The GMEE-WFED system, with its intricate layers and advanced modeling techniques, represents a significant leap forward in harnessing the potential of wind energy for a more sustainable future.

Analysis of solar energy potentials of five selected south-east cities in nigeria using deep learning algorithms

This study presents a meticulous examination of the solar energy potential of five selected metropolitan cities (Abakaliki, Awka, Enugu, Owerri, and Umuahia) in Eastern part of Nigeria using deep learning algorithm, specifically the Long Short-Term Memory (LSTM) model. These cities, despite being characterized by extended rainy seasons and a high level of cloudiness, are suitable environment for solar power generation and investment opportunities. The employed methodology capitalized on the LSTM deep learning approach to analyze and predict energy generation, utilizing comprehensive hourly weather data from the National Airspace Agency (NASA). The data set comprised various parameters, such as date/time, solar azimuth angle, temperature, humidity, wind speed, wind direction, cloud cover, and power, enabling a thorough analysis of each city. To ensure accuracy, energy prediction capabilities were benchmarked against real-time datasets from a solar power plant in Ulsan, South Korea, thereby training and fine-tuning the model for precision. The LSTM model's performance metrics were maintained at a learning rate of 0.07, a batch size of 150, and a train-test split ratio of 0.8 to 0.2. Data validation exhibited a mean square error (MSE) of 0.01, demonstrating the model’s reliability. Results showed Enugu as having the highest solar energy potential, averaging 6.25 kWh/day, while Awka registered the most substantial electricity demand across various sectors. These findings highlight the substantial potential for photovoltaic (PV) power systems and advocate for the immediate implementation of renewable energy policy in the selected cities. These are expected to bring about significant implications for future renewable energy environmentally friendly investments in Nigeria and globally.

The Predictability of the 30 October 2020 İzmir-Samos Tsunami Hydrodynamics and Enhancement of Its Early Warning Time by LSTM Deep Learning Network

Although tsunamis occur less frequently compared to some other natural disasters, they can be extremely devastating in the nearshore environment if they occur. An earthquake of magnitude 6.9 Mw occurred on 30 October 2020 at 12:51 p.m. UTC (2:51 p.m. GMT+03:00) and its epicenter was approximately 23 km south of İzmir province of Turkey, off the Greek island of Samos. The tsunami event triggered by this earthquake is known as the 30 October 2020 İzmir-Samos (Aegean) tsunami, and in this paper, we study the hydrodynamics of this tsunami using some of these artificial intelligence (AI) techniques applied to observational data. More specifically, we use the tsunami time series acquired from the UNESCO data portal at different stations of Bodrum, Syros, Kos, and Kos Marina. Then, we investigate the usage and shortcomings of the Long Short Term Memory (LSTM) DL technique for the prediction of the tsunami time series and its Fourier spectra. More specifically we study the predictability of the offshore water surface elevation dynamics, their spectral frequency and amplitude features, possible prediction success and enhancement of the accurate early prediction time scales. The uses and applicability of our findings and possible research directions are also discussed.

Anomaly Detection in Cyber Security with Graph-Based LSTM in Log Analysis

Intrusion detection systems utilize the analysis of log data to effectively detect anomalies. However, detecting anomalies quickly and effectively in large and heterogeneous log data can be challenging. To address this difficulty, this study proposes the GLSTM (Graph-based Long Short-Term Memory) framework, a graph-based deep learning model that analyzes log data to detect cyber-attacks rapidly and effectively. The framework involves standardizing the complex and diverse log data, training this data on an artificial intelligence model, and detecting anomalies. Initially, the complex and diverse log data is transformed into graph data using Node2Vec, enabling efficient and rapid analysis on the artificial intelligence model. Subsequently, these graph data are trained using LSTM (Long Short-Term Memory), Bi-LSTM, and GRU(Gated Recurrent Unit) deep learning algorithms. The proposed framework is tested using Hadoop’s HDFS dataset, collected from different systems and heterogeneous sources, as well as the BGL and IMDB datasets. Experimental results on the selected datasets demonstrate high levels of success.

Deep Learning Based Method for Early Warning of Price Risk in Railroad Bulk Freight Transport with Consideration of Time Series of Comprehensive Impact Index

The aim in this study is to enhance the price competitiveness of medium- and long-haul railroad bulk cargoes in the transportation market. Efficiently triggering an early warning for railroad bulk freight price risk is crucial. This paper discusses relevant factors affecting the pricing of railway freight from four aspects, namely the macroeconomy, the transportation market, the freight owner, and the railway enterprise. We construct a comprehensive index system for generating an early warning of railroad bulk freight prices. Utilizing a collected data set of 20 factor indicators that affected prices from 2015 to 2017, we calculate a comprehensive price risk warning index using the integrated entropy weight and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method. We propose a risk warning level classification method based on the k-means clustering method to determine five levels of price risk warning. Subsequently, we establish comprehensive index prediction methods for issuing price risk warnings based on the back propagation (BP) neural network and the deep learning long short-term memory (LSTM) network. We verify that the prediction performance of the deep learning LSTM network is significantly better than that of the BP neural network by combining all coal freight data from the Shanghai Railway Bureau for the period 2015 to 2017 as a case study. The proposed method can predict the level of freight price risk more accurately in the next phase, based on the status of the factor indicators, and be used to assist railway departments in developing a more reasonable freight price adjustment strategy to scientifically mitigate freight price risk.

Deep learning integration optimization of electric energy load forecasting and market price based on the ANN–LSTM–transformer method

Introduction: Power load forecasting and market price analysis have become crucial in the context of complex power energy systems and volatile market prices. Deep learning technology has gained significant attention in time series forecasting, and this article aims to enhance the accuracy and reliability of power load and market price predictions by integrating and optimizing deep learning models.Methods: We propose a deep learning framework that combines artificial neural networks (ANNs), long short-term memory (LSTM), and transformer models to address key challenges in electricity load forecasting and market price prediction. We leverage ANNs for their versatility and use LSTM networks for sequence modeling to generate initial predictions. Additionally, we introduce transformer technology and utilize its self-attention mechanism to capture long-distance dependencies within the data, further enhancing the model’s performance.Results: In our experiments, we validate the proposed framework using multiple public datasets. We compare our method with traditional forecasting approaches and a single-model approach. The results demonstrate that our approach outperforms other methods in predicting power load and market prices. This increased accuracy and reliability in forecasting can be of significant value to decision-makers in the energy sector.Discussion: The integration of deep learning models, including ANN, LSTM, and transformer, offers a powerful solution for addressing the challenges in power load and market price prediction. The ability to capture long-distance dependencies using the transformer's self-attention mechanism improves forecasting accuracy. This research contributes to the field of energy and finance by providing a more reliable framework for decision-makers to make informed choices in a complex and dynamic environment.

Aquaculture 4.0: hybrid neural network multivariate water quality parameters forecasting model

This study examined the efficiency of hybrid deep neural network and multivariate water quality forecasting model in aquaculture ecosystem. Accurate forecasting of critical water quality parameters can allow for timely identification of possible problem areas and enable decision-makers to take pre-emptive remedial actions that can significantly improve water quality management in aquaculture industry. A novel hybrid deep learning neural network multivariate water quality parameters forecasting model is developed with the aid of ensemble empirical mode decomposition (EEMD) method, deep learning long-short term memory (LSTM) neural network (NN), and multivariate linear regression (MLR) method. The presented water quality forecasting model (shortened as EEMD–MLR–LSTM NN model) is developed using multivariate time-series water quality sensor data collected from Loch Duart company, a Salmon offshore aquaculture farm based around Scourie, northwest Scotland. The performance of the novel hybrid water quality forecasting model is validated by comparing the forecast result with measured water quality parameters data and the real Phytoplankton data count from the aquaculture farm. The forecast accuracy of the results suggests that the novel hybrid water quality forecasting model can be used as a valuable support tool for water quality management in aquaculture industries.

Modelling and forecasting high-frequency data with jumps based on a hybrid nonparametric regression and LSTM model

High-frequency financial data is more difficult to predict than low-frequency data because it possesses nonlinearity, nonstationarity, higher volatility, and long memory and is frequently accompanied by the jump phenomena. In this paper, the nonparametric regression (NR) model based on kernel function is used to fit the nonlinear relationship between the nonstationary series Yt and its lagging series to model the trend of high frequency financial time series. Furthermore, the deep learning LSTM (long short-term memory) model is applied to capture the high volatility and frequent jumps of high frequency financial data and to improve the forecasting accuracy of the residual series. The results demonstrate that the hybrid NR and LSTM model has greatly improved the forecasting accuracy in several evaluation criteria. In comparison to NR, support vector machine (SVM), LSTM, ARIMA and NR-SVM models, the mean absolute error (MAE) of NR-LSTM has reduced by 89.78%, 97.85%, 86.48%, 32.47% and 89%, respectively. In addition, we have constructed the trading strategy for the Shanghai-Shenzhen 300 index by using the NR-LSTM model. The NR-LSTM model can continue to provide good returns even during a bear market, which can serve as a guide for investors. Furthermore, the NR-LSTM model also exhibits the best forecasting effect when we model the high-frequency data of Ping An bank in China, the FTSE 100 index in the UK, and the S&P 500 index in the US.

An interpretable LSTM deep learning model predicts the time-dependent swelling behavior in CERCER composite fuels

The study proposes a reliable and trustworthy data-driven approach with high precision and low computational cost, to predict the multiscale irradiation swelling behavior within CERCER composite fuels. A Long Short-Term Memory (LSTM) deep learning model is utilized to process the sequential data obtained from high-fidelity multiscale simulations. The model is evaluated using the metrics of accuracy, efficiency, interpretability, and generalizability. Compared to the feedforward and recurrent neural network models, the results indicate that the LSTM is superior at capturing the time-dependent character of the swelling progression and accurately catching the initiation of the critical recrystallization phenomena. The global and local interpretations via Shapley additive explanation theory (SHAP) agree with the mechanism analysis in simulations and experimental findings. The former identifies that the loading parameters of fission rate and density are the primary determinants in modulating the swelling deformation, while the latter uncovers the competing mechanism between temperature and hydrostatic pressure. The proposed model showcases its capacity to generalize on a time-consuming simulation involving a time-discrete algorithm. It guarantees a rapid swelling prediction in 36.4 secs that greatly reduces the time needed for 20.7 h in FE simulations.

High-precision polar motion prediction using EOP_20_C04 and EAM based on CSLS+AR and CSLS+LSTM methods

SUMMARY High-precision prediction of polar motion (PM) plays an important role in fields such as astronomy, geodesy, navigation and aerospace. Incorporating information on the effective angular momentum (EAM) of the geophysical fluid is an effective way to improve the precision of PM prediction. Based on the EOP_20_C04 data set and the EAM function, this study applies complex segmented least-squares (CSLS) + the long short term memory (LSTM) neural network and CSLS + autoregression (AR) models to predict PM. For the 6-d PM prediction, the mean absolute errors (MAEs) achieved by CSLS+AR are 1.03 and 0.8 mas in the X- and Y-directions, respectively, corresponding to reductions of 45.80 and 31.97 per cent when compared to predictions reported routinely in Bulletin A of the International Earth Rotation and Reference Systems Service (IERS). For the 365-d PM prediction, the MAEs gained by CSLS+LSTM model are 14.58 and 10.59 mas in the X- and Y-directions, respectively, corresponding to reductions of 28.17 and 51.09 per cent compared to predictions of the Bulletin A, and the prediction accuracy attained by CSLS+LSTM is higher than other prediction schemes. The experimental results show that, when considering EAM information, the CSLS+AR model can achieve better prediction accuracy for short-term PM prediction, however, the CSLS+LSTM model is more effective for medium- and long-term PM prediction owning to the excellent nonlinear fitting capability of the LSTM deep learning algorithm.

A Multilayer Approach for Intrusion Detection with Lightweight Multilayer Perceptron and LSTM Deep Learning Models

Intrusion detection is essential in the field of cybersecurity for protecting networks and computer systems from nefarious activity. We suggest a novel multilayer strategy that combines the strength of the Lightweight Multilayer Perceptron (MLP) and Long Short-Term Memory (LSTM) deep learning models in order to improve the precision and effectiveness of intrusion detection.The initial layer for extraction of features and representation is the Lightweight MLP. Its streamlined architecture allows for quick network data processing while still maintaining competitive performance. The LSTM deep learning model, which is excellent at identifying temporal correlations and patterns in sequential data, receives the extracted features after that.Our multilayer technique successfully manages the highly dimensional and dynamic nature of data from networks by merging these two models. We undertake extensive tests on benchmark datasets, and the outcomes show that our strategy performs better than conventional single-model intrusion detection techniques.The suggested multilayer method also demonstrates outstanding efficiency, which makes it particularly ideal for real-time intrusion detection in expansive network environments. Our multilayer approach offers a strong and dependable solution for identifying and reducing intrusions, strengthening the security position of computer systems and networks as cyber threats continue to advance.

LSTM Deep Learning Models for Virtual Sensing of Indoor Air Pollutants: A Feasible Alternative to Physical Sensors

Monitoring individual exposure to indoor air pollutants is crucial for human health and well-being. Due to the high spatiotemporal variations of indoor air pollutants, ubiquitous sensing is essential. However, the cost and maintenance associated with physical sensors make this currently infeasible. Consequently, this study investigates the feasibility of virtually sensing indoor air pollutants, such as particulate matter, volatile organic compounds (VOCs), and CO2, using a long short-term memory (LSTM) deep learning model. Several years of accumulated measurement data were employed to train the model, which predicts indoor air pollutant concentrations based on Building Management System (BMS) data (e.g., temperature, humidity, illumination, noise, motion, and window state) as well as meteorological and outdoor pollution data. A cross-validation scheme and hyperparameter optimization were utilized to determine the best model parameters and evaluate its performance using common evaluation metrics (R2, mean absolute error (MAE), root mean square error (RMSE)). The results demonstrate that the LSTM model can effectively replace physical indoor air pollutant sensors in the examined room, with evaluation metrics indicating a strong correlation in the testing set (MAE; CO2: 15.4 ppm, PM2.5: 0.3 μg/m3, VOC: 20.1 IAQI; R2; CO2: 0.47, PM2.5: 0.88, VOC:0.87). Additionally, the transferability of the model to other rooms was tested, with good results for CO2 and mixed results for VOC and particulate matter (MAE; CO2: 21.9 ppm, PM2.5: 0.3 μg/m3, VOC: 52.7 IAQI; R2; CO2: 0.45, PM2.5: 0.09, VOC:0.13). Despite these mixed results, they hint at the potential for a more broadly applicable approach to virtual sensing of indoor air pollutants, given the incorporation of more diverse datasets, thereby offering the potential for real-time occupant exposure monitoring and enhanced building operations.