Extended hierarchical extreme learning machine with multilayer perceptron
Khanittha Phurattanaprapin, Punyaphol Horata · 2016
For learning in big datasets, the classification performance of ELM might be low due to input samples are not extracted features properly. To address this problem, the hierarchical extreme learning machine (H-ELM) framework was proposed based on the hierarchical learning architecture of multilayer perceptron. H-ELM composes of two parts; the first is the unsupervised multilayer encoding part and the second part is the supervised feature classification part. H-ELM can give higher accuracy rate than of the traditional ELM. However, it still has to enhance its classification performance. Therefore, this paper proposes a new method namely as the extending hierarchical extreme learning machine (EH-ELM). For the extended supervisor part of EH-ELM, we have got an idea from the two-layers extreme learning machine. To evaluate the performance of EH-ELM, three different image datasets; Semeion, MNIST, and NORB, were studied. The experimental results show that EH-ELM achieves better performance than of H-ELM and the other multi-layer framework.