Agriculture Production Prediction of Major Fruits by Using Machine Learning Technique

Abstract

In the machine learning field, prediction of agricultural production is a challenging task. The main issue is providing guidance for the government and farmers, as crop production is less per year compared with the population, and the government provides facilities to farmers to increase their yield production. In this study, time series data from 1980–2015 were used. The data related to different types of fruits, namely apples, citrus, pears, grapes, and bananas, as well as total fruit production. The machine learning technique was used to predict the future production of ma fruits in Pakistan. The data were collected from the National Bureau of Statistics of Pakistan, and the output of major fruit production was 1000 tons. The linear regression technique was also used, and the results showed good accuracy. The findings suggested need to further increase the production of fruit; thus, the government needs to introduce new policies for farmers to improve the production of yield. This proposed method is applicable for developing countries (like third world countries) and they can compare production with increasing population and make some new policies to increase fruit production.