Evaluating the Robustness of Neural Network and ARIMA Models in Predicting Stock Prices: A Case Study of Tata Consultancy Services

Abstract

The present study focuses on evaluating the robustness of Artificial Neural Networks (ANNs) and Auto-Regressive Integrated Moving Average (ARIMA) models in forecasting stock prices. The analysis identifies a stochastic trend in the daily time series data from August 1, 2010, to August 1, 2024. The study period is characterized by high volatility, making this research distinct from existing literature. Empirical findings indicate that ARIMA is the most suitable model for this dataset. The selection of the best model is based on Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). Although both models yield closely comparable results, ARIMA emerges as the superior choice for the data under consideration.