Design and Analysis of Multiclass Classification Models for Rainfall Prediction
Department of Computer Science and Engineering, Birla Institute of Technology Mesra, Panta campus, Bihar, India.
The extraction of useful information from large amount of data using data mining is today’s demand. Extraction is based on common behaviour of a real world process by observing the trends and relationships of the data elements. Weather forecasting, rainfall forecasting and event detection such as storm or cyclone prediction play an important role in day to day applications. In this paper, we propose a methodology for rainfall prediction. The proposed methodology involves the application of typical K-Means clustering on the weather database of Patna. The rainfall category of new coming data is predicted by predicting the cluster number or the label of the data point. This model can efficiently predict the class to which a certain data point will belong to and hence can categorize the type of rainfall. Categorization of rainfall done into four labels and related model has been simulated using MLlib and R programming.
Key words: K-Means Clustering, Classification, Decision Tree, Random Forest, Multiple Linear Regression, Backward Elimination