Experimenting with Extreme Learning Machine for Biomedical Image Classification

Currently, deep learning networks, with particular regard to convolutional neural network models, are typically exploited for biomedical image classification. One of the disadvantages is that extremely expensive train due complex data models. Extreme machine has recently emerged which, as shown in experimental studies, can produce an acceptable predictive performance several classification tasks, and at a much lower training cost compared networks trained by backpropagation. We propose method devoted exploring possibility considering extreme machines tasks. Binary multiclass four case studies considered demonstrate effectiveness machine, images acquired dermatoscope blood cell microscope, showing be successfully applied