A COMBINATION OF TEMPORAL SEQUENCE LEARNING AND DATA DESCRIPTION FOR ANOMALYBASED NIDS

Abnormal identification of oil monitoring based on LSTM and SVDD

Forecasting daily PM2.5 concentrations in Wuhan with a spatial-autocorrelation-based long short-term memory model

Abstract: This research is driven by the challenges faced by Universitas Lancang Kuning (UNILAK) in attracting applicants amidst intense competition, especially after the government's policy opened independent pathways to State Universities (PTN) from 2022-2023, which impacted private university applicant numbers. To address this and support strategic planning, this study aims to predict the trend of prospective students applying to all study programs at UNILAK for the period 2025-2027. Two time series models were employed: ARIMA (AutoRegressive Integrated Moving Average) and LSTM (Long Short-Term Memory). Applicant data from 2019 to 2024 was used to build the model. The Augmented Dickey-Fuller (ADF) test confirmed the data's stationarity with a p-value of 0.0. ACF and PACF analyses determined the ARIMA parameters as p=1, d=1, q=1. The LSTM model was trained to capture more complex data patterns. ARIMA predictions for 2025, 2026, and 2027 are 3298.66, 3362.33, and 3371.30, respectively. LSTM predictions for the same years are 3335.64, 3476.52, and 3518.42. Evaluation using Root Mean Squared Error (RMSE) showed ARIMA (RMSE=588.72) to be more accurate than LSTM (RMSE=653.96). Nevertheless, LSTM provided a more optimistic prediction. This study concludes that ARIMA is better suited for short-term planning, while LSTM can be used for more ambitious long-term strategies. Keywords: arima; LSTM; applicants; prediction; university Abstrak: Penelitian ini didorong oleh tantangan Universitas Lancang Kuning (UNILAK) dalam menarik pendaftar di tengah persaingan ketat, khususnya setelah kebijakan pemerintah membuka jalur mandiri ke Perguruan Tinggi Negeri (PTN) sejak 2022-2023, yang menyebabkan penurunan jumlah pendaftar di universitas swasta. Untuk mendukung perencanaan strategis, studi ini bertujuan memprediksi tren jumlah calon mahasiswa yang mendaftar ke seluruh program studi di UNILAK untuk periode 2025-2027.Dua model deret waktu digunakan: ARIMA (AutoRegressive Integrated Moving Average) dan LSTM (Long Short-Term Memory). Data jumlah pendaftar dari 2019 hingga 2024 digunakan untuk membangun model. Uji Augmented Dickey-Fuller (ADF) menunjukkan data stasioner dengan p-value 0,0. Analisis ACF dan PACF menentukan parameter ARIMA sebagai p=1, d=1, q=1. Model LSTM dilatih untuk menangkap pola data yang lebih kompleks.Prediksi ARIMA untuk 2025, 2026, dan 2027 adalah 3298.66, 3362.33, dan 3371.30. Prediksi LSTM untuk tahun yang sama adalah 3335.64, 3476.52, dan 3518.42. Evaluasi menggunakan Root Mean Squared Error (RMSE) menunjukkan ARIMA (RMSE=588.72) lebih akurat daripada LSTM (RMSE=653.96). Meskipun demikian, LSTM memberikan prediksi yang lebih optimis. Studi ini menyimpulkan ARIMA lebih cocok untuk perencanaan jangka pendek, sementara LSTM dapat digunakan untuk strategi jangka panjang yang ambisius. Kata kunci: arima; LSTM; pendaftar; prediksi; universitas

Widespread use of internet connected devices results in a flow of data among the connected devices. This has led to creating these network systems prone to different type of cyber attacks among which network intrusion is one of a kind. To detect network intrusion, an intrusion detection system (IDS) is placed in the system. The previous works in this field classify the attack mainly with traditional machine learning algorithm based models. This experiment proposes Long Short Term Memory (LSTM) based neural network model along with two other machine learning models i.e. Support Vector Machine (SVM), K-Nearest Neighbor (KNN), which shows significant improved performance in intrusion detection. The performance of the model is proved to be effective comparing to the previous works and also deployable in real life environments which is validated by using explainable AI. For this experiment, Kyoto University Honeypot log dataset has been used to build and evaluate the performances of the IDS model.KeywordsIntrusion detectionLSTMSVMLIMEClass imbalance

Long short-term memory - Fully connected (LSTM-FC) neural network for PM2.5 concentration prediction

Traffic safety has been thought of as a basic feature of transportation, recent developments in civil aviation have emphasized the need for risk identification and safety prediction. This study aims to increase en-route flight safety through the development of prediction models for flight conflicts. Firstly, flight conflicts time series and traffic parameters are extracted from historical ADS-B data. In the second step, a Long Short-Term Memory (LSTM) model is trained to make a one-step-ahead prediction on the flight conflict time series. The results show that the LSTM model has the greatest prediction effect (MAE 0.3901) with comparison to other models. Based on that, we add traffic parameters (volume, density, velocity) into the LSTM model as new input variables and issue a comprehensive analysis of the relative predictive power of traffic parameters. The accuracy of prediction model is validated with a mean error of less than 3%. Based on the improvements of model performance brought by traffic parameters, LSTM models with a single traffic parameter are proposed for further discussion. The results illustrate that volume is the most important factor in promoting prediction accuracy and density has an advantage of improvement in the aspect of model stability.

The accurate prediction of carbon dioxide (CO2) emissions in the building industry can provide data support and theoretical insights for sustainable development. This study proposes a hybrid model for predicting CO2 emissions that combines a multi-strategy improved particle swarm optimization (MSPSO) algorithm with a long short-term memory (LSTM) model. Firstly, the particle swarm optimization (PSO) algorithm is enhanced by combining tent chaotic mapping, mutation for the least-fit particles, and a random perturbation strategy. Subsequently, the performance of the MSPSO algorithm is evaluated using a set of 23 internationally recognized test functions. Finally, the predictive performance of the MSPSO-LSTM hybrid model is assessed using data from the building industry in the Yangtze River Delta region as a case study. The results indicate that the coefficient of determination (R2) of the model reaches 0.9677, which is more than 10% higher than that of BP, LSTM, and CNN non-hybrid models and demonstrates significant advantages over PSO-LSTM, GWO-LSTM, and WOA-LSTM hybrid models. Additionally, the mean square error (MSE) of the model is 2445.6866 Mt, and the mean absolute error (MAE) is 4.1010 Mt, both significantly lower than those of the BP, LSTM, and CNN non-hybrid models. Overall, the MSPSO-LSTM hybrid model demonstrates high predictive accuracy for CO2 emissions in the building industry, offering robust support for the sustainable development of the industry.

SummaryFor the past few years, Bitcoin plays a vital role in both the economical and financial industries. In order to gain a huge return on investment, the investors are eager to forecast the future value of Bitcoin. However, Bitcoin price variation is quite nonlinear and chaotic in nature, so it creates more difficulty in forecasting future value. Researchers found that the multiplicative long short term memory (LSTM) model will be more efficient for predicting those complex variations. So, target mission is about to develop an optimized multiplicative LSTM with an Attention mechanism using Technical Indicators derived from historical data. A modified cuckoo search optimization model is proposed to tune the hyperparameter of the Deep Learning model. This novel optimization algorithm eliminates the local optimum and slower convergence problem of the cuckoo search optimization algorithm. Deibold Mariano test is performed to statistically evaluate the proposed model and it is inferred that the recommended methodology is statistically fit. Regression metrics such as root mean square error, mean square error and mean absolute error has been used for comparative evaluation with related benchmark techniques such as genetic algorithm optimized LSTM (GA–LSTM), particle swarm optimized LSTM (PSO–LSTM) and cuckoo search optimized LSTM (CSO–LSTM). The empirical result shows that the recommended methodology outperforms the taken benchmark models and provides better accuracy.

Urban air quality index forecasting using multivariate convolutional neural network based customized stacked long short-term memory model

With the rapid global proliferation of electric vehicles (EVs), their integration as a significant load component within power systems increasingly influences the stable operation and planning of electrical grids. However, the high uncertainty and randomness inherent in EV users’ charging behaviors render accurate load forecasting a challenging task. In this context, the present study proposes a Particle Swarm Optimization (PSO)-enhanced Long Short-Term Memory (LSTM) network forecasting model. By combining the global search capability of the PSO algorithm with the advantages of LSTM networks in time-series modeling, a PSO-LSTM hybrid framework optimized for seasonal variations is developed. The results confirm that the PSO-LSTM model effectively captures seasonal load variations, providing a high-precision, adaptive solution for dynamic grid scheduling and charging infrastructure planning. This model supports the optimization of power resource allocation and the enhancement of energy storage efficiency. Specifically, during winter, the Mean Absolute Error (MAE) is 3.896, a reduction of 6.57% compared to the LSTM model and 10.13% compared to the Gated Recurrent Unit (GRU) model. During the winter–spring transition, the MAE is 3.806, which is 6.03% lower than that of the LSTM model and 12.81% lower than that of the GRU model. In the spring, the MAE is 3.910, showing a 2.71% improvement over the LSTM model and a 7.32% reduction compared to the GRU model.

Most Graph Neural Networks (GNNs) follow the message-passing scheme. Residual connection is an effective strategy to tackle GNNs' over-smoothing issue and performance reduction issue on non-homophilic networks. Unfortunately, the coarse-grained residual connection still suffers from class-imbalanced over-smoothing issue, due to the fixed and linear combination of topology and attribute in node representation learning. To make the combination flexible to capture complicated relationship, this paper reveals that the residual connection needs to be node-dependent, layer-dependent, and related to both topology and attribute. To alleviate the difficulty in specifying complicated relationship, this paper presents a novel perspective on GNNs, i.e., the representations of one node in different layers can be seen as a sequence of states. From this perspective, existing residual connections are not flexible enough for sequence modeling. Therefore, a novel node-dependent residual connection, i.e., Long Short-Term Graph Memory Network (LSTGM) is proposed to employ Long Short-Term Memory (LSTM), to model the sequence of node representation. To make the graph topology fully employed, LSTGM innovatively enhances the updated memory and three gates with graph topology. A speedup version is also proposed for effective training. Experimental evaluations on real-world datasets demonstrate their effectiveness in preventing over-smoothing issue and handling networks with heterophily.

Network intrusion poses a severe threat to the Internet. Intrusion detection methods based on deep learning are very effective to process and analyze intrusion data. On the one hand, they use the Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) models to extract spatial and temporal features, respectively. However, they either adopt a single model or operate two models in series. And it fails to capture temporal-spatial features effectively. On the other hand, previous methods do not consider the multi-feature correlation of intrusion data. And then they cannot get better classification performance. To address the two above problems, we propose a hybrid neural network (HNN) model by integrating multi-feature correlation and temporal-spatial analysis. First, we adopt a contribution-based feature selection. Second, we reconstruct multi-feature correlation and then apply the CNN and LSTM in parallel to extract temporal-spatial features. Finally, we splice the temporal-spatial features with the correlation features, or study the influence of the temporal-spatial features by attention mechanism. Based on the above operations, we exploit the Deep Neural Network (DNN) to detect intrusion data. The experimental results show that HNN improves 3.78%, 1.31%, 0.21%, and 1.13% accuracy on the UNSW-NB15, AWID, CICIDS 2017, and CICIDS 2018 datasets.

Sensitivity analysis of isoprene and aerosol emission in a suburban plantation using long short-term memory model

High-resolution prediction of the spatial distribution of PM2.5 concentrations in China using a long short-term memory model

This study develops a ship propeller diagnostic system to address the issue of insufficient ship maintenance capacity and enhance operational efficiency. It uses the Remaining Useful Life (RUL) prediction technology to establish a sensing platform for ship propellers to capture vibration signals during ship operations. The Diagnosis and RUL Prediction Model is designed to assess bearing aging status and the RUL of the propeller. The synchronized signal reshaping technology is employed in the Diagnosis and RUL Prediction Model to process the original vibration signals as input to the model. The vibration signals obtained are used to analyze the temporal and spectral energy of propeller bearings. Exponential functions are used to generate the health index as model outputs. Model inputs and outputs are simultaneously input into a Long Short-Term Memory (LSTM) model for training, culminating as the Diagnosis and RUL Prediction Model. Compared to Recurrent Neural Network and Support Vector Regression models used in previous studies, the Diagnosis and RUL Prediction Model developed in this study achieves a Mean Squared Error (MSE) of 0.018 and a Mean Absolute Error (MAE) of 0.039, demonstrating outstanding performance in prediction results and computational efficiency. This study integrates the Diagnosis and RUL Prediction Model, bearing aging experimental data, and real-world vibration measurements to develop the diagnosis and RUL prediction system for ship propellers. Experiments with ship propellers show that when the bearing of the propeller enters the worn stage, this diagnostic system for ship propellers can accurately determine the current status of the bearing and its remaining useful life. This study offers a practical solution to insufficient ship maintenance capacity and contributes to improving the operational efficiency of ships.