Comparison of Extreme Learning Machines and Support Vector Machines on Premium and Regular Gasoline Classification for Arson and Oil Spill Investigation.

摘要

Extreme Learning Machine (ELM) is a recently introduced learning algorithm for single hiddenlayer feedforward neural network. Compared with classical learning algorithms in neural network, e.g. Back Propagation, ELM can achieve better performance with much shorter learning time. In the existing literature, its better performance and comparison with Support Vector Machine (SVM), over regression and general classification problems catch the attention of many researchers. In this paper, the comparison between ELM and SVM over a particular area of classification, i.e. gasoline classification for arson and oil spill investigation, is conducted. Detection and correct identification of gasoline types during arson and fuel spill investigation are very important in forensic science. As the number of arson and oil spillage increases, it becomes very important to have an accurate means of detecting and classifying gasoline found at such sites of incidence. However, currently only a very few number of classification models have been explore in this germane field of forensic science, particularly for gasoline identification. Comparison of simulation results for SVM and ELM show that, for the different categories of gasoline classification investigated, SVM still outperforms ELM in terms of percentage of correctly classified gasoline while in term of time taken for both training and testing, ELM clearly outperform SVM.