Unsupervised Difficulty Estimation with Action Scores

Luis Octavio Arriaga Camargo, Matias Valdenegro-Toro

In: Workshop: LXAI Research @ NeurIPS 2020. LatinX in AI Research Workshop (LXAI-2020) befindet sich NeurIPS 2020 December 7 Virtual United States arXiv 2020.


Evaluating difficulty and biases in machine learning models has become of extreme importance as current models are now being applied in real-world situations. In this paper we present a simple method for calculating a difficulty score based on the accumulation of losses for each sample during training. We call this the action score. Our proposed method does not require any modification of the model neither any external supervision, as it can be implemented as callback that gathers information from the training process. We test and analyze our approach in two different settings: image classification, and object detection, and we show that in both settings the action score can provide insights about model and dataset biases.

Weitere Links

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence