Jun 2026· The Arabian journal for science and engineering· 0 citations· 15 references
TL;DR
The developed scheme outperforms counterpart models, advocating its potential to enhance dynamic traffic management and intelligent signal coordination systems and enables more accurate short-term flow estimates, thereby reducing average vehicle waiting times and improving intersection-level signal responsiveness.
Abstract
In smart mobility networks, accurate vehicular flow forecasting is of critical importance, enabling efficient, robust, user- and environment-friendly management of devices, technologies, and systems. However, current short-term traffic prediction algorithms frequently face challenges of computational inefficiency and limited predictive precision. To overcome these limitations, this paper introduces an LSTM model optimized through Bayesian optimizer (BO-LSTM) to enhance prediction accuracy. The traffic data is first preprocessed through data augmentation using random sampling and scaling of traffic counts, which is particularly suitable for traffic time series data as it preserves temporal patterns while increasing data diversity and robustness against demand fluctuations. After this augmentation step, the data is standardized to bring the input features to a similar range. The LSTM model is trained using Bayesian optimization for hyperparameters tuning, including the learning rate, dense layers, number of iterations, and dropout rate, within an acceptable range. The root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE) are computed for three different datasets, measured in number of vehicles per minute, yielding respective values of 0.0442, 0.0353, and 2.91%, 0.0216, 0.0173, and 1.84%, and 0.4426, 0.3541, and 2.63%. These results correspond to improvements of 34.5, 59.5, and 32.6% in RMSE, respectively, when compared against Attention-LSTM and temporal convolutional network (TCN) models. This proves the accuracy of the proposed model. The developed scheme outperforms counterpart models, advocating its potential to enhance dynamic traffic management and intelligent signal coordination systems. Such capability enables more accurate short-term flow estimates, thereby reducing average vehicle waiting times and improving intersection-level signal responsiveness.
Accurate Traffic Flow Forecasting (TFF) is important for emerging Intelligent Transportation Systems (ITS) that support active traffic management, optimize routes, and reduce congestion. In this paper, Deep Learning (DL) methods for TFF, with an emphasis on models like Recurrent Neural Networks (RNN) reinforced with attention mechanism, Bidirectional Long Short-Term Memory (Bi-LSTM), as well as Stacked Autoencoder (SAE) is used for ITS. In complex traffic situations, these strategies improve prediction accuracy and remove nonlinear spatial-temporal networks. Bio-inspired optimization methods, such as the Fruit Fly Optimization Algorithm (FFOA), Philippine Eagle Optimization (PEO) and Kookaburra Optimization Algorithm (KOA) are reviewed for adaptive learning, weight initialization and optimal parameter adjustment in order to further improve model performance. The model architectures, optimization techniques, and assessment criteria discussed in recent research are compared in this review to show how they contribute to precise RMSE, MAPE, MAE traffic forecasts. With a focus on multi-source data fusion, real-time adaptability and interpretable AI frameworks for next-generation ITS, it concludes by identifying research gaps and future creativities.
V. Poornima, M. Subashini· International Conference on...· 0 citations
Nonlinear fluctuations and abrupt events in traffic flow can lead to significant point prediction errors, thereby limiting the effectiveness of traffic control strategies in Intelligent Transportation Systems (ITS). To address this issue, we propose a novel hybrid deep learning model, Bayes-BP-Transformer. This model combines Bayesian inference with backpropagation and a Transformer architecture, improving prediction accuracy while maintaining the model's generalization ability. Specifically, the backpropagation component captures local temporal patterns, while the Transformer encoder models long-range spatiotemporal dependencies. By introducing Bayesian inference, the model optimizes the probability distribution of network weights, improving robustness to noise and incomplete data. Experiments on two real-world traffic datasets demonstrate the superior performance of our model. Compared to advanced models, Bayes-BP-Transformer reduces the root mean square error (RMSE) by 18.6-38.0% and the mean absolute percentage error (MAPE) by 5.7-32
A. Feofilova, V. Fialkin, Jixiao Jiang· The eurasia proceedings of s...· 0 citations
Accurate short-term traffic prediction in Long-Term Evolution (LTE) networks is essential for proactive resource management and maintaining quality of service. With increasing data demand and dynamic traffic patterns, reliable forecasting at the cell level has become critical for efficient network operations. This analysis compared ARIMA, SARIMA, LSTM and GRU models with real hourly traffic data of urban and rural LTE eNodeBs in Nepal. The dataset was split chronologically into 80% training and 20% testing before preprocessing to prevent data leakage. The LSTM model performed best, achieving MSE of 1.4804, MAE of 0.8770, RMSE of 1.2167, and R-squared of 0.9563. Deep learning, particularly LSTM, was able to learn the non-linear and complex traffic patterns better than other models. This finding highlights the practical applicability of deep learning models in real world telecom network operations. The research provides a comparative analysis of the statistical and deep learning models for real-time, cell-level LTE eNodeB traffic forecasting across diverse geospatial environments using real-world data, with practical validation through LSTM- based traffic prediction for network optimization.
This study presents a comprehensive review of machine learning (ML) and deep learning (DL) techniques for traffic flow prediction, identifying key methodological trends, performance strengths, and existing research gaps. The review systematically examines ML approaches such as K-means LSTM, KNN, and SVM, highlighting their effectiveness in short-term forecasting and feature-driven prediction scenarios. In addition, advanced DL architectures including CNN–LSTM, GRGCAN, and TS-RNN are analysed for their ability to capture complex spatial–temporal dependencies in traffic data, demonstrating superior adaptability and predictive accuracy in dynamic traffic environments. A key contribution of this study lies in its comparative synthesis of ML and DL models, revealing that hybrid and graph-based DL architectures consistently outperform traditional ML methods when handling large-scale, heterogeneous traffic datasets. The review further identifies critical limitations in existing studies, particularly challenges related to real-time deployment, limited integration of external factors such as weather and traffic incidents, and data quality constraints. Based on these findings, the study highlights the importance of incorporating multimodal transport data and external contextual information to improve prediction robustness. Overall, this review provides actionable insights for researchers and practitioners, supporting the development of more reliable and adaptive intelligent transportation systems to reduce urban congestion and improve traffic management.
Thabo Matue, A. A. Akinyelu, Mase Mokotsolane· International Journal of Dat...· 0 citations
Long short-term memory (LSTM) networks, a class of recurrent neural networks (RNNs), are widely used for sequential data modeling and time series forecasting. Their predictive accuracy, however, strongly depends on architectural choices and hyperparameter settings. In this study, we formulate LSTM architecture optimization as a mixed-variable combinatorial problem and propose a hybrid framework combining Particle Swarm Optimization (PSO) with the Adam optimizer to adaptively tune various LSTM variants, including standard LSTM, stacked LSTM (sLSTM), multiplicative LSTM (mLSTM), and extended LSTM (xLSTM). The approach is evaluated on two traffic flow datasets: the Metro Interstate dataset with daily forecasting intervals, and the PeMS datasets with one-hour ahead predictions across multiple sensor networks. Experimental results reveal a consistent performance hierarchy: the PSO-optimized xLSTM achieves the highest predictive accuracy (MAE = 257.83, RMSE = 385.97, SMAPE = 11.88%, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^2 = 0.961$$\end{document} on Metro Interstate; MAE = 6.056, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R^2 = 0.992$$\end{document} on PeMS08), followed by the mLSTM, while standard LSTM and sLSTM lag behind. These findings demonstrate that framing LSTM architecture optimization as a mixed-variable combinatorial problem, coupled with PSO-based optimization, substantially improves forecasting performance, offering a robust and versatile strategy for accurate traffic prediction and other complex sequential data applications.
Taoufyq Elansari, Hamza H. Sulimani· Evolutionary Systematics· 0 citations
Efficient traffic-state prediction at urban intersections is a critical component of intelligent transportation systems(ITS), as traffic conditions are influenced by dynamic factors such as traffic demand variability, infrastructure constraints, and operational traffic-control policies. This study proposes a deep-learning-based approach for short-term traffic-state classification using real-world traffic data collected during 2022 at the Alésia intersection in Paris. The objective is to classify traffic conditions into five operational states: Unknown, Flowing, Pre-saturated, Saturated, and Blocked. To investigate the impact of temporal modeling on traffic-state recognition, four deep learning architectures were evaluated under identical experimental conditions: Artificial Neural Networks (ANN), Simple Recurrent Neural Networks (RNN),Long Short-Term Memory networks (LSTM), and Gated Recurrent Units (GRU). Considering the highly imbalanced nature of the dataset, model performance was assessed using complementary metrics including Accuracy, Precision, Recall, F1-score, Macro-F1 score, and Balanced Accuracy. Experimental results demonstrate that recurrent architectures substantially outperform the ANN baseline, highlighting the importance of temporal dependencies in traffic-state classification. While the conventional RNN achieves high overall accuracy, its performance on minority traffic states remains limited. Among the evaluated models, the LSTM achieves the highest Balanced Accuracy (70.91%), indicating superior recognition of underrepresented traffic conditions. The GRU attains the highest overall F1-score (0.9256) and Macro-F1 score (0.497), while maintaining competitive classification accuracy (91.01%), providing the most favorable trade-off between global predictive performance and balanced class-wise recognition.The analysis of learning curves, classification reports, and confusion matrices further confirms the effectiveness of gated recurrent architectures for handling highly imbalanced multiclass traffic-state classification problems. These findings provide practical insights for the deployment of intelligent traffic-monitoring systems capable of supporting real-time traffic management and decision-making in urban environments.
Chaymae Chouiekh, Ali Yahyaouy, M. A. Sabri et al.· Vehicles· 0 citations