Extended performance appraise of Bayes, Function, Lazy, Rule, Tree data mining classifier in novel transformed fractional content based image classification

Extended performance appraise of Bayes, Function, Lazy, Rule, Tree data mining classifier in novel transformed fractional content based image classification Image classification has become one of the important research field as hundreds of images are generated everyday which implies the need to build the classification system. To build faster and easy classification system, the visual content of images is used. Accuracy of classification depends upon the feature extraction which is one of the most important step in image classification. The paper shows the performance of additional four orthogonal transforms using transformed fractional content as feature for image classificationwhere the Kekre, Hartle, Slant and Haar transform are used in addition to earlier proposed use of sine, cosine and walsh transforms. Twelve assorted classifiers across five datamining classifier family (Bayes, Function, Lazy, Rule and Tree) are used. Here 504 number of variations for proposed image classification method are experimented using twelve classifiers, seven orthogonal transforms and six fractions of transformed content. The Simple Logistic classifiers with Kekre transform gives better image classification closely followed by Simple Logistic with sine transform and Simple Logistic with Hartley transform.