Fuzzy Association Rules for Prediction of Monsoon Rainfall

A fuzzy association rule algorithm is implemented to extract the relationship between the atmospheric indices and the Indian Summer Monsoon Rainfall (ISMR). ENSO and EQWIN indices are used as the causative variables. Rules extracted are showing a negative relationship with ENSO index and a positive relationship with the EQWIN index. A fuzzy rule based prediction technique is also implemented on the same indices in order to predict the ISMR. Rules are defined using a training dataset for the period 1958 – 1999 and validated for the period 2000 – 2006. The fuzzy outputs of the defined rules are converted into crisp outputs using the weighted counting algorithm. The variability of the ISMR over the years is captured by this technique, thus proving to be efficient even when the linear statistical relationship between the indices is weak.