arXiv cs.AI / cs.LG / cs.CL·18d agoThe Surprising Effectiveness of Approximate Value Iteration in Self-Play#alphazero#approximate-value-iteration#game-playingAI research