REAL TIME TRAFFIC FLOW PREDICTION AND ANOMALY DETECTION USING BIG DATA

This paper represented a real-time traffic flow prediction and anomaly detection framework leveraging big data analytics and advanced ML models. The combination of spatio-temporal deep learning techniques with streaming anomaly detection provides accurate forecasting and reliable abnormal event identification. Future work includes incorporating probabilistic forecasting to quantify uncertainty, applying transfer learning for cross-city model adaptation, integrating multimodal transportation data, and deploying lightweight compressed models for edge computing environments. Further research may also explore reinforcement learning-based traffic control strategies using predicted traffic states and detected anomalies.

[1]    J. Lv et al., “Traffic Flow Prediction With Big Data: A Deep Learning Approach,” IEEE Transactions on Intelligent Transportation Systems, 2015. [2]      Y. Li et al., “Diffusion Convolutional Recurrent Neural Network for Traffic Forecasting,” International Conference on Learning Representations, 2018. [3]      B. Yu et al., “Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting,” IJCAI, 2018. [4]    W. Ma et al., “LSTM Neural Network for Traffic Speed Prediction Using Remote Microwave Sensor Data,” Transportation Research Part C, 2015. [5]    Z. Cui et al., “Traffic Graph Convolutional Recurrent Neural Network,” IEEE Access, 2019. [6]      X. Jia et al., “Prediction-Based Traffic Anomaly Detection Using Big Data Analytics,” IEEE Access, 2020. [7]      H. Zheng et al., “Urban Traffic Flow Prediction: A Deep Learning Approach,” IEEE Transactions on Intelligent Transportation Systems, 2019. [8]     Y. Wu et al., “Graph WaveNet for Deep Spatial-Temporal Graph Modeling,” IJCAI, 2019. [9]      L. Sun et al., “Using Autoencoders for Traffic Anomaly Detection,” IEEE Intelligent Transportation Systems Conference, 2017. [10]      M. Zaharia et al., “Apache Spark: A Unified Engine for Big Data Processing,” Communications of the ACM, 2016.