Classification of Categorical and Numerical Data on Selected Subset of Features
Zaher Al Aghbari · BiblioBoard Library Catalog (Open Research Library) · 2010
In this paper, we presented a wrapper approach to select the best subset of features that result in the highest classification accuracy. We use an SFS approach to search for the best subset of features. The Naïve Bayes algorithm and K-Nearest Neighbor algorithm are used to classify and estimate the accuracy of the categorical data and image data, respectively. This approach is evaluated using two datasets: categorical teachers’ dataset and image dataset. The experimental results for both categorical and image datasets show the feasibility of the presented techniques in classifying categorical and numerical data. Such techniques are useful in many applications to decrease the performance cost and increase the classification accuracy.