Uncertainty-based Rejection Wrappers for Black-box Classifiers
| Autor: | Oriol Pujol, José Mena, Jordi Vitrià |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Artificial intelligence
General Computer Science Computer science 02 engineering and technology 010501 environmental sciences Machine learning computer.software_genre 01 natural sciences Robustness (computer science) Aprenentatge automàtic 0202 electrical engineering electronic engineering information engineering General Materials Science uncertainty Categorical variable 0105 earth and related environmental sciences business.industry Intel·ligència artificial General Engineering Probabilistic logic Classification rejection systems Sistemes classificadors (Intel·ligència artificial) Data point machine learning 020201 artificial intelligence & image processing lcsh:Electrical engineering. Electronics. Nuclear engineering business computer lcsh:TK1-9971 Learning classifier systems |
| Zdroj: | Dipòsit Digital de la UB Universidad de Barcelona IEEE Access, Vol 8, Pp 101721-101746 (2020) |
| Popis: | Machine Learning as a Service platform is a very sensible choice for practitioners that want to incorporate machine learning to their products while reducing times and costs. However, to benefit their advantages, a method for assessing their performance when applied to a target application is needed. In this work, we present a robust uncertainty-based method for evaluating the performance of both probabilistic and categorical classification black-box models, in particular APIs, that enriches the predictions obtained with an uncertainty score. This uncertainty score enables the detection of inputs with very confident but erroneous predictions while protecting against out of distribution data points when deploying the model in a productive setting. We validate the proposal in different natural language processing and computer vision scenarios. Moreover, taking advantage of the computed uncertainty score, we show that one can significantly increase the robustness and performance of the resulting classification system by rejecting uncertain predictions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|