[2509.04959] On the Normalization of Confusion Matrices: Methods and Geometric Interpretations

Computer Science > Machine Learning arXiv:2509.04959 (cs) [Submitted on 5 Sep 2025 (v1), last revised 30 Mar 2026 (this version, v2)] Title:On the Normalization of Confusion Matrices: Methods and Geometric Interpretations Authors:Johan Erbani, Pierre-Edouard Portier, Elod Egyed-Zsigmond, Sonia Ben Mokhtar, Diana Nurbakova View a PDF of the paper titled On the Normalization of Confusion Matrices: Methods and Geometric Interpretations, by Johan Erbani and 4 other authors View PDF HTML (experimental) Abstract:The confusion matrix is a standard tool for evaluating classifiers by providing insights into class-level errors. In heterogeneous settings, its values are shaped by two main factors: class similarity -- how easily the model confuses two classes -- and distribution bias, arising from skewed distributions in the training and test sets. However, confusion matrix values reflect a mix of both factors, making it difficult to disentangle their individual contributions. To address this, we introduce bistochastic normalization using Iterative Proportional Fitting, a generalization of row and column normalization. Unlike standard normalizations, this method recovers the underlying structure of class similarity. By disentangling error sources, it enables more accurate diagnosis of model behavior and supports more targeted improvements. We also show a correspondence between confusion matrix normalizations and the model's internal class representations. Both standard and bistochasti...