Price forecasting of Brinjal: A statistical evaluation
Author(s):
Pal Vikash, Darji VB and Chaudhari Raju
Abstract:
A timely and reliable forecast of prices for different agricultural crops is highly needed and required for now a day. Forecasting of prices for agricultural commodities remains difficult because they are influenced by many factors. The uncertainty of future price, production and consumption level makes agricultural market strategy and investment planning difficult. Perishability, price changes and seasonal nature of vegetables affect a lot to the vegetable prices. In the present study linear, quadratic and exponential trends were used for trend studies and forecasting purpose. Also ARIMA, ARCH/GARCH models and Artificial Neural Network (ANN) were employed for the study. For the studying error behaviour Jarque-Bera test was utilized. Statistical comparisons were made for different models using Root Mean Squared Error (RMSE) and Mean Absolute Per cent Error (MAPE). Jarque-Bera test results showed that none of the model residuals followed normal distribution. In all, the comparison of different models tried in the study to forecast prices for Brinjal the Artificial Neural Network (ANN) model on the basis of RMSE value performed better as compared to all the models studied.
How to cite this article:
Pal Vikash, Darji VB and Chaudhari Raju. Price forecasting of Brinjal: A statistical evaluation. The Pharma Innovation Journal. 2021; 10(12S): 2105-2110.