Weather forecasting is done by collecting huge data with high quantity and quality from the earth atmosphere by scientifically understanding about the atmospheric changes. The dimensionality of the data is increasing day to day. At the same time, it has become a difficult task to identify the relevant features to apply the data-mining algorithm more effectively. Feature selection helps to extract the relevant number of features from the huge amount of data. It faces several limitations when the feature selection is done in a big data set as input. This paper introduces a regression analysis approach in the feature selection scenario. The input is taken from Atlantic hurricane database. It uses principal component regression analysis for effectively performing feature selection and the linear regression is performed to extract the more significant features that would improve the effectiveness and efficiency of the data. In future, this Principle component regression technique can be used to perform the clustering accuracy.
Volume 11 | 04-Special Issue
Pages: 2403-2408