Closed Item-Set Mining for Prediction of Indian Summer Monsoon Rainfall A Data Mining Model with Land and Ocean Variables as Predictors

Practical application of data mining in scientific and engineering domains, when explored, pose many problems and provide interesting results. In this paper, we attempt to mine out association rules from 37 (1969–2005) years of Indian summer monsoon rainfall data and try its applicability in helping better prediction of Indian summer monsoon rainfall. We shortlist 36 variables as possible predictors of Indian summer monsoon rainfall based on previous literature and compare prediction using all 36 variables and prediction by selected attributes from derived association rules. Results show better performance in prediction of All India region, West central region and Peninsular region rainfall when attributes selection is employed as compared to all 36 variables used for prediction.