Breast Cancer Classification Using AdaBoost- Extreme Learning Machine

Abstract

Extreme learning machine (ELM) is a new learning algorithm with single-hidden layer feed-forward neural network (SLFN). In this study, an ensemble of ELMs is used to predict breast cancers. To this end, an AdaBoost-based algorithm is proposed to adapt ELM in ensemble learning. A threshold value is defined for reweighting part of data that is misclassified. Based on the threshold-value, a tuning parameter is used. This parameter is defined as the inverse of the number of neurons in ELM's hidden-layer. Then, an experiment is conducted for tuning the number of neuron in hidden-layer. The results indicate that the proposed ensemble learning can effectively improves the recall and precision of the classification.