An Efficient Approach for Rainfall Forecasting Using Data Mining

Farming as the major occupation of India requires a scientific and analytic focus for its productivity. Agricul-ture and allied sectors like forestry and fisheries accounted for 15-19% of the GDP (Gross Domestic Product) [1]. India is a country having various seasons and different geographical conditions. India’s climate can be classified as a hot tropical country, except the northern states of Himachal Pradesh and Jammu and Kashmir in north and Sikkim in the north eastern hills, which have a cooler, more continental influenced climate. These variations in climate make prediction of weather difficult. Techniques like Machine Learning algorithms are required for studying and predicting the weather conditions. This paper represents an analysis and prediction of rainfall by studying the previous data accumulated in 100 of years [2]. This paper uses basic techniques of Data Mining to conduct a trend analysis on Rainfall Data. We have analyzed data of various regions, implemented suitable data mining techniques based on literature survey to achieve our goal of analyzing and predicting the Rainfall. Hierarchical clustering is used for grouping and clustering similar data together. Regression technique is used to predict the range of values, Graph Extrapolation technique is used to estimate value and to predict the future pattern.