Comparative Analysis of Data Mining Models for Crop Yield by Using Rainfall and Soil Attributes

Till the seventies of the last century, Indian agriculture was in a poor condition. The agrarian economy was largely consumption-oriented and there were poor irrigation facilities and simple agricultural implements. Agricultural yield was very low and dependency on nature was very high. The food grains were not enough to feed the population. With a view to augment the yield, the Indian government had no option but to introduce Green Revolution. The Green Revolution was a movement towards excessive mechanisation of agriculture. The agriculturists were motivated and assisted to undertake the technology-based farming. Irrigation facilities were developed. However, the results of Green Revolution were not uniform all over the country. Neither has there has been uniform impact on all kinds of crops nor has there been uniform impact on all the regions and all categories of farmers. Even today the farmer falls prey to the risks unleashed by the nature. Success or failure of rain fed vegetation depends upon the sample and amounts of rainfall. But, other factors like temperature, photoperiod and grid additionally notably influence crop boom and yield. The analysis of climate performs a key role in planning better farming structures to enhance and stabilise yields, and to design appropriate crop breeding strategies. With the use of technology, it has also become possible to minimise the risks involved in agriculture to which the early farmers were awfully exposed. There are in particular two procedures to predict rainfall. Empirical technique and dynamical method. In our method we use the empirical technique that is based on evaluation of historical information of the rainfall and its dating to a spread of atmospheric variables over different components of the nation. The most broadly used empirical approaches used for weather prediction are regression, artificial neural network, fuzzy logic and institution approach of statistics dealing with. We use data mining techniques such as clustering and classification techniques for rainfall prediction.