AN INTEGRATED APPROACH COMBINING NEURAL NETWORKS AND GENETIC ALGORITHMS FOR MULTISOURCE TIME SERIES FORECASTING

Abstract

This paper proposes a multisource data-driven approach to enhance time series forecasting in the tourism domain. Specifically, we integrate a multilayer perceptron (MLP) neural network with the NSGA-II genetic algorithm to optimize the model’s hyperparameters, replacing manual tuning or traditional GA-based methods. In addition to historical data on monthly international tourist arrivals to Vietnam, two exogenous data sources are incorporated: (i) sentiment indices extracted from tourist reviews on TripAdvisor, and (ii) search trend data from Google Trends. Key hyperparameters such as input window size, learning rate, number of epochs, early stopping, and normalization methods are optimized through the genetic algorithm. Experimental results indicate that the NSGA-II-MLP model using multisource data outperforms the standard MLP and maintains stable performance across both pre-COVID-19 and pandemic periods. The results underscore the effectiveness of combining multisource data with NSGA-II optimization for tourism demand forecasting.