Learningtosearchefficientlyforcausally near-optimaltreatments
–Neural Information Processing Systems
Tosatisfythenear -optimalityconstraintof(1), weusean estimateofthefunction (h), see(2), todefine , , (h):=1[ (h)< / ]forparameters , 0, 1. Thefunction , , (h)representswhetheran , / -optimumhasbeenfound. Wedefine r , , (h, a)=
Neural Information Processing Systems
Feb-7-2026, 11:45:54 GMT