K-Means and Competitive Neural Network Based Clustering of Climate Data

The city Nainital is a beautiful hill station of Uttarakhand, India. The summers of this city is very pleasant, as compared to other places, due to its very favorable climatic conditions. The climatic secondary data of Nainital is the plot of this research work. This data is comprised of 90 samples containing the values of minimum temperature, maximum temperature and rainfall for three seasons, namely summer, winter and rainy. The aim of this work is to design clustering models, on basis of this data, for the exact classification of inputs into three classes (i.e. summer, winter and rainy). The two different clustering techniques, adopted in this work, using MATLAB, are K-means clustering and the neural network based competitive learning approach. Finally, the clustering performance of both the techniques is compared.