Classification Using the Compact Rule Generation

Various attributes within a dataset relate to each other and with the class attribute. The relationship between the different attributes with class attribute may improve the classification accuracy. The paper introduces CCSA algorithm that performs the clustering that is cascaded by classification based on association. The Clustering process generates a group of various instances within the dataset. These clustered instances are classified by using the association. This paper uses the Apriori association to generate the rules for classification. The technique is analyzed by using the soil data set and various other online available datasets using WEKA. The simulation result using the WEKA shows that reduced rules with the improved classification accuracy as compared to the existing association with classification algorithms.