Study of the effect of attributes' selection on the quality of the classification models created with algorithms available in R language.

The aim of the project is to investigate the effect of attribute selection type on the quality of classification models by using the algorithms available in R language.

Classification algorithms available in R require the indication of a set of attributes on which classification has to find a way of mapping the data into a set of predefined classes. This project will examine the impact of the selection of these attributes on the quality of the classification model.

There will be a matrix of data (read from .csv file) and a parameter specifying the number of attributes that will have to be selected by the studied selection algorithms. After selecting attributes the program will build the classification model based on data-training using one of the classification algorithms available in R language. Then algorithm will test the accuracy of the model using other available data and compare the results with the results obtained using other search strategy.

The quality of the selection method will be evaluated based on the quality of the created classification model. This quality is determined by the ratio of accuracy - the percentage of test examples correctly classified by the model.