From historical data to future predictions: Analyzing and forecasting oilseed yield trends in India using time series methods

Abstract

The study investigates the use of Autoregressive Integrated Moving Average (ARIMA) modeling techniques to predict oilseed yields using historical agricultural data. The dataset includes records of oilseed yields from multiple growing seasons in India. The study uses data preprocessing, cleaning, exploratory analysis, and rigorous stationarity checks. The ARIMA model's parameters are identified through Autocorrelation Function and Partial Autocorrelation Function plots. Performance is evaluated using Akaike Information Criterion corrected, Bayesian Information Criterion, Root Mean Squared Error, and residual analysis. The findings show the ARIMA model's effectiveness in capturing temporal patterns and seasonality in oilseed yield data, proving its ability to provide accurate forecasts.