Active Learning Is Planning: Nonmyopic Ε-Bayes-Optimal Active Learning of Gaussian Processes
Abstract
A fundamental issue in active learning of Gaussian processes is that of the exploration-exploitation trade-off. This paper presents a novel nonmyopic ε-Bayes-optimal active learning ( ε -BAL) approach [4] that jointly optimizes the trade-off. In contrast, existing works have primarily developed greedy algorithms or performed exploration and exploitation separately. To perform active learning in real time, we then propose an anytime algorithm [4] based on ε -BAL with performance guarantee and empirically demonstrate using a real-world dataset that, with limited budget, it outperforms the state-of-the-art algorithms.
Cite
Text
Hoang et al. "Active Learning Is Planning: Nonmyopic Ε-Bayes-Optimal Active Learning of Gaussian Processes." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2014. doi:10.1007/978-3-662-44845-8_43Markdown
[Hoang et al. "Active Learning Is Planning: Nonmyopic Ε-Bayes-Optimal Active Learning of Gaussian Processes." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2014.](https://mlanthology.org/ecmlpkdd/2014/hoang2014ecmlpkdd-active/) doi:10.1007/978-3-662-44845-8_43BibTeX
@inproceedings{hoang2014ecmlpkdd-active,
title = {{Active Learning Is Planning: Nonmyopic Ε-Bayes-Optimal Active Learning of Gaussian Processes}},
author = {Hoang, Trong Nghia and Low, Kian Hsiang and Jaillet, Patrick and Kankanhalli, Mohan S.},
booktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases},
year = {2014},
pages = {494-498},
doi = {10.1007/978-3-662-44845-8_43},
url = {https://mlanthology.org/ecmlpkdd/2014/hoang2014ecmlpkdd-active/}
}