Statistical Inference of the Matthews Correlation Coefficient for Multiclass Classification
Jun Tamura, Yuki Itaya, Kenichi Hayashi, Kouji Yamamoto · 2025
Read Statistical Inference of the Matthews Correlation Coefficient for Multiclass Classification by Jun Tamura, Yuki Itaya, Kenichi Hayashi, Kouji Yamamoto (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-06450.txt.
Description
Jun Tamura and co-authors set out a clear framework for a balanced score used to judge classification models when there are three or more classes, in a paper from 2025. The score is valued because it stays fair when classes are uneven in size. Yet its multi class forms were used without firm definitions.
The team defines two averaged forms of the score and gives methods to build confidence ranges for them. They also extend the methods to compare two scores in paired studies. Tests with simulated and real data showed the methods work well. The work aims to stop users from trusting a single point value without a sense of its reliability.
Catalog record
| CREATOR | Jun Tamura, Yuki Itaya, Kenichi Hayashi, Kouji Yamamoto |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 38 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-06450