Online POMDP Planning with Anytime Deterministic Guarantees - Supplementary Anonymous Author(s) Affiliation Address email

W e restate some of the definitions from the paper for convenience. Moreover, depending on the algorithm implementation, the number of iterations can be finite (e.g. by After the allowable time steps ended, the simulation was reset to its initial state. The main goal is to tag as many opponents as possible within a given time frame. The states reflect the agent's current position and the opponents' positions. The Baby POMDP is a classic problem that represents the scenario of a baby and a caregiver. The states in this problem represent the baby's needs, which could be hunger, The observations are binary, either the baby is crying or not.