Extreme learning machine for mammographie risk analysis

Abstract

The assessment of mammographie risk analysis is an important issue in the medical field. Various approaches have been applied in order to achieve a higher accuracy in such analysis. In this paper, an approach known as Extreme Learning Machines (ELM), is employed to generate a single hidden layer neural network based classifier for estimating mammographie risk. ELM is able to avoid problems such as local minima, improper learning rate, and overfitting which iterative learning methods tend to suffer from. In addition the training phase of ELM is very fast. The performance of the ELM-trained neural network is compared with a number of state of the art classifiers. The results indicate that the use of ELM entails better classification accuracy for the problem of mammographie risk analysis.