Wilt Dataset-based Comparative Analysis of Three Neural Networks

This paper carries out a comparative analysis of three neural networks and their specific kernel machine learning algorithms, namely Inverse Weighted K-means in Online mode (IWKO1), evolving Fuzzy Optimally Pruned Extreme Learning Machine (eF-OP-ELM) and Support Vector Machine (SVM). The presentation of IWKO1, eF-OP-ELM and SVM is given and associated with their mathematical background, and the steps of their implementations are next outlined. The comparative analysis involves two performance indices, prediction accuracy and execution time, measured after the implementation on the Wilt dataset. The analysis is important in the context of the need to handle big data volumes in terms of using special algorithms, which are specially designed to process appropriately such data and information.