Exploring the Effectiveness of Neural Network Algorithms on Mobility Patterns: A Comprehensive Metrics-Based Study.

Abstract

Accurate location prediction in mobility models ($\mathbf{x}$ and y coordinates) is vital in transportation planning, urban development, and optimizing mobile networks. This study investigates the efficacy of various neural network algorithms, including Multilayer Perceptron (MLP), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), and Long ShortTerm Memory (LSTM), in analyzing and predicting mobility patterns. Utilizing a metrics-based approach, we assess these models across different synthetic mobility models, employing Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE) as evaluation criteria. Our findings offer valuable insights into the strengths and limitations of each neural network architecture in the context of mobility pattern analysis, contributing to the optimization of future predictive models in this domain.