The nexus between food price inflation and monsoon rainfall in India: exploring through comparative data mining models

Abstract

In the paper we analysed the impact on food inflation of monsoon rainfall using different data mining tools, namely: lda (Linear Discriminant Analysis), qda (Quadratic Discriminant Analysis), lr (logistic regression), rpart (Recursive Partitioning and Regression Trees), knn (k-Nearest Neighbors Network), and formulated the models in such a way that food inflation at the end of the financial year can be predicted from the rainfall received during the monsoon month of the year, and a few other known variables. The study is expected to be useful as it can predict the chances of high food inflation with 65% and 63% of accuracy by rpart and lr models, respectively. This information on the chances of high food inflation just after monsoon months can be very useful for policy-makers. While prediction of high food inflation will not in itself solve the problem, it would help decision-makers to take precautionary measures to minimize its adverse impacts on the population.