Average Performance Analysis of Multi-Class Classification Based on Error-Correcting Output Codes
Manabu Kobayashi, Gendo Kumoi, Hideki Yagi, Shigeichi Hirasawa · 2024
In machine learning, one of the methods to solve multiclass classification problems is a framework called Error-Correcting Output Codes (ECOC), which constructs a multiclass classifier by combining a lot of binary classifiers. ECOC assigns binary codewords to each category, and the multiclass classification performance varies depending on the code. In this study, we treat each element of the codeword as a random variable and evaluate the average performance of ECOC. As a result, for$M$class classification if the number of binary classifiers is$O(\log M)$, then the average error probability of various codes approaches that of MAP estimation. We show that the important points are the ratio between the number of binary classifiers and$\log M$and the difference between the maximum posterior probability and the second highest posterior probability for the categories.