FORECASTING OF AREA AND PRODUCTION OF CASHEW NUT IN DAKSHINA KANNADA USING ARIMA AND EXPONENTIAL SMOOTHING MODELS

  • Chaithra M., Department of Agricultural Statistics, Applied Mathematics and Computer Science, University of Agricultural Sciences, Bengaluru, India
  • Pramit Pandit Department of Agricultural Statistics, Applied Mathematics and Computer Science, University of Agricultural Sciences, Bengaluru, India
  • Bishvajit Bakshi Department of Agricultural Statistics, Applied Mathematics and Computer Science, University of Agricultural Sciences, Bengaluru, India
Keywords: ARIMA, Cashew Nut, Exponential Smoothing, Forecasting, Time Series Analysis

Abstract

The cultivation and marketing of cashew nut involve a considerable amount of work force. Hence, it plays a vital role in the Indian economic scenario. In this context, an attempt has been made to forecast the area and production of cashew nut with a view to help the planners in recommending policies regarding cashew nut. Due to autocorrelation in the data, time series forecasting models such as ARIMA and exponential smoothing models were adopted. Detection and removal of 3 significant outliers, i.e. 1 for area under cashew nut and 2 in case of cashew nut production, were done before fitting the models. Holt’s model was found to have better forecasting ability with lowest RMSE value (1386.13) among the different models fitted for forecasting the area under cashew nut. From this model, area (ha) under cashew nut was forecasted to be 34492.10, 34974.81 and 35474.87 for the year 2018, 2019 and 2020, respectively. In case of cashew nut production, Brown’s linear trend model, with RMSE value (10020.19), was observed to have better forecasting ability among the tried models. Production of cashew nut (in tonnes) was forecasted to be 10230.20, 10996.81 and 11833.00 for the year 2018, 2019 and 2020, respectively.

 

Published
2019-09-30
Section
Articles