Gene Expression Data Using Extreme Learning Machine

The aim of this work is to develop a new technique for medical classification problem. In diagnosing cancer, multicategory classification of cancer plays a very significant role. Nowadays, number of cancer suffering is increasing, so effective technique is required. In this work, an Extreme Learning Machine is integrated with the Successive Feature Selection technique for better classification in cancer. The extreme learning machine will rectify the problems such as improper learning rate, local minima, and low speed. This new method is evaluated for using accuracy and execution time and proves that the proposed method is very effective.