We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Entropy methods for the confidence assessment of probabilistic classification models

Abstract: Many classification models produce a probability distribution as the outcome of a prediction. This information is generally compressed down to the single class with the highest associated probability. In this paper, we argue that part of the information that is discarded in this process can be in fact used to further evaluate the goodness of models, and in particular the confidence with which each prediction is made. As an application of the ideas presented in this paper, we provide a theoretical explanation of a confidence degradation phenomenon observed in the complement approach to the (Bernoulli) Naive Bayes generative model.
Comments: 12 pages, 7 figures
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
Cite as: arXiv:2103.15157 [stat.ML]
  (or arXiv:2103.15157v1 [stat.ML] for this version)

Submission history

From: Gabriele Tornetta [view email]
[v1] Sun, 28 Mar 2021 15:39:13 GMT (95kb)

Link back to: arXiv, form interface, contact.