Abstract
We study the problem of efficient PAC active learning of homogeneous linear classifiers (halfspaces) in Rd, where the goal is to learn a halfspace with low error using as few label queries as possible. Under the extra assumption that there is a t-sparse halfspace that performs well on the data (t ≪ d), we would like our active learning algorithm to be attribute efficient, i.e. to have label requirements sublinear in d. In this paper, we provide a computationally efficient algorithm that achieves this goal. Under certain distributional assumptions on the data, our algorithm achieves a label complexity of O(t · polylog(d, 1 ∊ )). In contrast, existing algorithms in this setting are either computationally inefficient, or subject to label requirements polynomial in d or 1 ∊ .
Original language | English (US) |
---|---|
Pages (from-to) | 1856-1880 |
Number of pages | 25 |
Journal | Proceedings of Machine Learning Research |
Volume | 75 |
State | Published - 2018 |
Externally published | Yes |
Event | 31st Annual Conference on Learning Theory, COLT 2018 - Stockholm, Sweden Duration: Jul 6 2018 → Jul 9 2018 |
ASJC Scopus subject areas
- Artificial Intelligence
- Software
- Control and Systems Engineering
- Statistics and Probability