collaborators

6 papers

cs.LG2024

Markov flow policy -- deep MC

Nitsan Soffair, Gilad Katz

Discounted algorithms often encounter evaluation errors due to their reliance on short-term estimations, which can impede their efficacy in addressing simple, short-term tasks and…

cs.AI2024

Optimizing Agent Collaboration through Heuristic Multi-Agent Planning

Nitsan Soffair

The SOTA algorithms for addressing QDec-POMDP issues, QDec-FP and QDec-FPS, are unable to effectively tackle problems that involve different types of sensing agents. We propose a n…

cs.LG2024

Solving Collaborative Dec-POMDPs with Deep Reinforcement Learning Heuristics

Nitsan Soffair

WQMIX, QMIX, QTRAN, and VDN are SOTA algorithms for Dec-POMDP. All of them cannot solve complex agents' cooperation domains. We give an algorithm to solve such problems. In the fir…

cs.AI2024

Conservative DDPG -- Pessimistic RL without Ensemble

Nitsan Soffair, Shie Mannor

DDPG is hindered by the overestimation bias problem, wherein its -estimates tend to overstate the actual -values. Traditional solutions to this bias involve ensemble-based me…

cs.LG2024

MinMaxMin -learning

Nitsan Soffair, Shie Mannor

MinMaxMin -learning is a novel optimistic Actor-Critic algorithm that addresses the problem of overestimation bias (-estimations are overestimating the real -values) inher…

cs.LG2024

SQT -- std -target

Nitsan Soffair, Dotan Di-Castro, Orly Avner +1

Std -target is a conservative, actor-critic, ensemble, -learning-based algorithm, which is based on a single key -formula: -networks standard deviation, which is an "un…