Performance Analysis of the Regression and Time Series Predictive Models using Parallel Implementation for Agricultural Data

Abstract

The profit and loss of any type of business depends on the effective decision making and predictive models playvery important role in making wise decisions in all cases. Agriculture is also a kind business, where the farmers need to make the timely decisions in crop selection to make the farm business more useful and profitable. In developing countries like India, farmers are failing to make theright decision at right time to make the profitable farm business. An efficient predictive model based system could definitely help the farmers in making appropriate timely decisions in crop selection and thereby minimizing the loss. Hence there is a need for an efficient predictive model that can effectively deal with Big-data. As there are many predictive models existing, the performance of these models needs to be analyzed to select the more appropriate and effective model based on the application with the increasing data size. This paper proposes and implements a parallel predictive model for potentially forecasting the demand, supply and price values of the agricultural commodities. The forecasted values of various models are compared with the actual market values and analyzed the performance to select the best.