Experimental Analysis of Agricultural Data Using Data Mining

In the agriculture field farmers and agribusinesses have to make numerous decisions every day and complexities involve the various factors impacting them. Data mining techniques are the approach necessary for fulfilling effective solutions for this problem. Data Mining is a technique that focuses on large datasets to extract information for the prediction and discovery of hidden patterns. Agriculture mainly depends on climate or weather, agricultural topography, etc. The current study presents the numerous data mining/refining techniques and their role in the context of soil fertility, nutrient analysis. A decision tree is a good way for classification in data mining. C4.5, Classification and Regression Trees (ID3) are two mostly used decision tree algorithms for classification. ID3 algorithm produces misclassification, the main drawback of the C4.5 algorithm is that errors when the domain of the target attribute is very large. This paper therefore presents a modified decision tree algorithm to reduce limitations. The model is tested with the data set of soil samples. The test proves that the modified decision tree algorithm has higher classification accuracy when compared to C4.5 and ID3 algorithms. Classification of soil is that the separation of soil into classes or groups each having similar characteristics and potentially similar behavior. Classification of soil is needed so that farmers can know the type of soil and can plough the crops depending on the type of soil.