eng
Schloss Dagstuhl – Leibniz-Zentrum für Informatik
Leibniz International Proceedings in Informatics
1868-8969
2020-01-06
37:1
37:22
10.4230/LIPIcs.ITCS.2020.37
article
Learning and Testing Variable Partitions
Bogdanov, Andrej
1
https://orcid.org/0000-0002-0338-6151
Wang, Baoxiang
2
https://orcid.org/0000-0002-2997-0970
Department of Computer Science and Engineering , Institute of Theoretical Computer Science and Communications, The Chinese University of Hong Kong
Department of Computer Science and Engineering, The Chinese University of Hong Kong
Let F be a multivariate function from a product set Σ^n to an Abelian group G. A k-partition of F with cost δ is a partition of the set of variables V into k non-empty subsets (X_1, ̇s, X_k) such that F(V) is δ-close to F_1(X_1)+ ̇s+F_k(X_k) for some F_1, ̇s, F_k with respect to a given error metric. We study algorithms for agnostically learning k partitions and testing k-partitionability over various groups and error metrics given query access to F. In particular we show that
1) Given a function that has a k-partition of cost δ, a partition of cost O(k n^2)(δ + ε) can be learned in time Õ(n^2 poly 1/ε) for any ε > 0. In contrast, for k = 2 and n = 3 learning a partition of cost δ + ε is NP-hard.
2) When F is real-valued and the error metric is the 2-norm, a 2-partition of cost √(δ^2 + ε) can be learned in time Õ(n^5/ε^2).
3) When F is Z_q-valued and the error metric is Hamming weight, k-partitionability is testable with one-sided error and O(kn^3/ε) non-adaptive queries. We also show that even two-sided testers require Ω(n) queries when k = 2.
This work was motivated by reinforcement learning control tasks in which the set of control variables can be partitioned. The partitioning reduces the task into multiple lower-dimensional ones that are relatively easier to learn. Our second algorithm empirically increases the scores attained over previous heuristic partitioning methods applied in this context.
https://drops.dagstuhl.de/storage/00lipics/lipics-vol151-itcs2020/LIPIcs.ITCS.2020.37/LIPIcs.ITCS.2020.37.pdf
partitioning
agnostic learning
property testing
sublinear-time algorithms
hypergraph cut
reinforcement learning