Hybrid CNN-LSTM Models for Traffic Flow Prediction in Smart Cities
Srinivas Chalasani, S Ramya
In this research, we explore the effectiveness of hybrid CNN-LSTM models for traffic flow prediction
in smart cities, comparing their performance against traditional statistical methods, classical machine
learning algorithms, and standalone deep learning models. Traffic flow prediction is critical for
optimizing urban transportation systems, yet it poses significant challenges due to the complex and
dynamic nature of urban traffic patterns. The proposed hybrid CNN-LSTM model leverages the spatial
feature extraction capabilities of Convolutional Neural Networks (CNNs) and the temporal sequence
modeling strength of Long Short-Term Memory (LSTM) networks. Experimental results demonstrate
that the hybrid model outperforms all other methods, achieving the lowest Mean Absolute Error (MAE)
and Root Mean Squared Error (RMSE), alongside the highest R² score. These findings highlight the
potential of hybrid CNN-LSTM architectures in enhancing real-time traffic management and smart city
planning.
