A Comparative Study of Classical and Machine Learning Approaches in Time Series Prediction

Harpinder Kaur

Abstract: Time Series Analysis (TSA) is an essential analytical approach which operates across extensive areas. The purpose of this approach is to analyse historical data patterns to predict future values. Time series forecasting uses an analytical method that predicts future outcomes from data points collected over extended time periods. The research investigates various intricate forecasting methods, including Vector Auto Regression (VAR) and ARIMA, among others. Sophisticated time series models enable organizations to make better decisions while streamlining their processes and predicting future trends. The research illustrates recent developments in time series forecasting methods that emphasize real - time analysis and accuracy enhancement through external data integration, while merging traditional techniques with machine learning algorithms.

Keywords: TSA, ARIMA, VAR, FIR, and SIR