An experimental comparison between ELM and C4.5 for classification problems with symbolic attributes

It is well known that feed-forward neural networks can be learnt from symbolic data although the learnt networks usually have poor performance. This paper explores the ability of a recently popular feed-forward neural network, i.e., Extreme Learning Machine (ELM) for modeling symbolic data. An experimental study is conducted to compare C4.5 (a very popular algorithm of learning from symbolic data) with ELM by using ten categorical data sets. The comparison involves three aspects, i.e., learning speed, testing accuracy and impact of training data numbers on testing accuracy. Some comparative results and comments which may provide new insights to ELM community are given.