Multiclass Classification of Brain Cancer with Multiple Multiclass Artificial Bee Colony Feature Selection and Support Vector Machine

Abstract

A World Health Organization reported that the mortality rate due to brain cancer is the highest in the Asian continent. It is critical importance that brain cancer can be detected earlier so that the treatment process can be carried out more precisely and will be able to extend the life expectancy of brain cancer patients. Taking advantage of microarray data, machine learning methods can be applied to help brain cancer prediction according to its type. This problem can be referred to as a multiclass classification problem. Using the one versus one approach, there will be as many as two-class problems, where k indicates the number of classes. In this paper, Multiple Multiclass Artificial Bee Colony (MMABC) implemented as a feature selection method and Support Vector Machine (SVM) as a classification method. ABC algorithm proved successful in solving optimisation problems with high dimensionality, and SVM can produce accurate and robust classification results. The data obtained from Broad Institute data. The data consist of 7129 features and 42 samples. From the experiment, the accuracy of Multiple SVM using a feature selection based MMABC method reached 95.24% accuracy in usage 300 best features; this percentage slightly more superior than SVM method without feature selection.