Exploratory machine learning with unknown unknowns

DOI10.1016/J.ARTINT.2023.104059arXiv2002.01605OpenAlexW4389956001MaRDI QIDQ6152674FDOQ6152674

Authors: Peng Zhao, Jia-Wei Shan, Yujie Zhang, Zhi-Hua Zhou 

Publication date: 13 February 2024

Published in: Artificial Intelligence (Search for Journal in Brave)

Abstract: In conventional supervised learning, a training dataset is given with ground-truth labels from a known label set, and the learned model will classify unseen instances to the known labels. In this paper, we study a new problem setting in which there are unknown classes in the training dataset misperceived as other labels, and thus their existence appears unknown from the given supervision. We attribute the unknown unknowns to the fact that the training dataset is badly advised by the incompletely perceived label space due to the insufficient feature information. To this end, we propose the exploratory machine learning, which examines and investigates the training dataset by actively augmenting the feature space to discover potentially unknown labels. Our approach consists of three ingredients including rejection model, feature acquisition, and model cascade. The effectiveness is validated on both synthetic and real datasets.

Full work available at URL: https://arxiv.org/abs/2002.01605

Recommendations

• Generalized mixture models, semi-supervised learning, and unknown class inference
• Unsupervised supervised learning. I: Estimating classification and regression errors without labels
• Learning a classifier when the labeling is known
• Anomaly and novelty detection for robust semi-supervised learning
• Semi-supervised novelty detection

zbMATH Keywords

robustnessunknown unknownsexploratory machine learningrobust AI

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05)

Cites Work

• Bagging predictors
• Action elimination and stopping conditions for the multi-armed bandit and reinforcement learning problems
• A survey on concept drift adaptation
• Ensembling neural networks: Many could be better than all
• Classification with reject option
• Advanced Lectures on Machine Learning
• On optimum recognition error and reject tradeoff
• Reinforcement learning. An introduction
• Classification with a reject option using a hinge loss
• Handwritten digit recognition by combined classifiers.
• Active learning
• Mixed rational soliton solutions of two differential-difference equations in Casorati determinant form
• Classification methods with reject option based on convex risk minimization
• Improving classification performance through selective instance completion
• Learning with Rejection
• On reject and refine options in multicategory classification
• Foundations of machine learning
• Fast classification rates without standard margin assumptions

This page was built for publication: Exploratory machine learning with unknown unknowns

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6152674)