Efficient Planning in Large MDPs with Weak Linear Function Approximation
Roshan Shariff & Csaba Szepesvári
–Neural Information Processing Systems
To achieveour result, westartfrom theapproximatelinear programming(ALP) approach where the value function is approximated using the feature vectors.
Neural Information Processing Systems
Feb-19-2026, 07:59:11 GMT
- Country:
- Technology: