Confidence estimation of classification based on the distribution of the neural network output layer

TL;DR

This paper introduces novel confidence estimation methods for neural network classifiers that leverage the output logit distribution, improving prediction filtering without altering existing models.

Contribution

The authors propose new confidence estimation techniques based on logit distributions that require no modifications to neural networks and enhance prediction accuracy across various tasks.

Findings

Significant accuracy improvements in relation extraction, named entity recognition, and image classification.

Methods effectively filter predictions to maximize precision with minimal recall loss.

No changes needed to existing neural network architectures.

Abstract

One of the most common problems preventing the application of prediction models in the real world is lack of generalization: The accuracy of models, measured in the benchmark does repeat itself on future data, e.g. in the settings of real business. There is relatively little methods exist that estimate the confidence of prediction models. In this paper, we propose novel methods that, given a neural network classification model, estimate uncertainty of particular predictions generated by this model. Furthermore, we propose a method that, given a model and a confidence level, calculates a threshold that separates prediction generated by this model into two subsets, one of them meets the given confidence level. In contrast to other methods, the proposed methods do not require any changes on existing neural networks, because they simply build on the output logit layer of a common neural network. In particular, the methods infer the confidence of a particular prediction based on the distribution of the logit values corresponding to this prediction. The proposed methods constitute a tool that is recommended for filtering predictions in the process of knowledge extraction, e.g. based on web scrapping, where predictions subsets are identified that maximize the precision on cost of the recall, which is less important due to the availability of data. The method has been tested on different tasks including relation extraction, named entity recognition and image classification to show the significant increase of accuracy achieved.