pith. sign in

arxiv: 1202.5298 · v2 · pith:UYJCRZCMnew · submitted 2012-02-23 · 💻 cs.SY · cs.LG· cs.SY

Min Max Generalization for Two-stage Deterministic Batch Mode Reinforcement Learning: Relaxation Schemes

classification 💻 cs.SY cs.LGcs.SY
keywords relaxationproblemschemesbatchconstraintsdeterministicfirstlearning
0
0 comments X
read the original abstract

We study the minmax optimization problem introduced in [22] for computing policies for batch mode reinforcement learning in a deterministic setting. First, we show that this problem is NP-hard. In the two-stage case, we provide two relaxation schemes. The first relaxation scheme works by dropping some constraints in order to obtain a problem that is solvable in polynomial time. The second relaxation scheme, based on a Lagrangian relaxation where all constraints are dualized, leads to a conic quadratic programming problem. We also theoretically prove and empirically illustrate that both relaxation schemes provide better results than those given in [22].

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.