1. Trang chủ
  2. » Tất cả

trí tuệ nhân tạothan lambert,inst eecs berkeley edu

42 3 0
Tài liệu đã được kiểm tra trùng lặp

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 42
Dung lượng 1,44 MB

Nội dung

trí tuệ nhân tạothan lambert,inst eecs berkeley edu Game Playing State of the Art http //bit ly/3GEMok9 Chinook beat 40 year reign of champion Marion Tinsley using complete 8 piece endgame 2007 Checke[.]

Game Playing State-of-the-Art http://bit.ly/3GEMok9 CuuDuongThanCong.com Chinook beat 40-year-reign of champion Marion Tinsley using complete 8-piece endgame 2007: Checkers solved! Chess: 1997: Deep Blue defeats human champion Gary Kasparov in a six-game match Deep Blue examined 200M positions per second, used sophisticated evaluation and undisclosed methods for extending some lines of search up to 40 ply Current programs are even better, if less historic Go: Human champions are being beaten In go, b > 300! Classically use pattern knowledge bases, but big recent advances use Monte Carlo (randomized) expansion methods Pacman https://fb.com/tailieudientucntt Behavior from Computation http://bit.ly/3GEMok9 CuuDuongThanCong.com Demo : mysterypacman(L6D1) https://fb.com/tailieudientucntt Video of Demo Mystery Pacman CuuDuongThanCong.com https://fb.com/tailieudientucntt Adversarial Games http://bit.ly/3GEMok9 CuuDuongThanCong.com https://fb.com/tailieudientucntt Types of Games http://bit.ly/3GEMok9 CuuDuongThanCong.com Many different kinds of games! Axes: t Deterministic or stochastic? t One, two, or more players? t Zero sum? t Perfect information (can you see the state)? Want algorithms for calculating a strategy (policy) which recommends a move from each state https://fb.com/tailieudientucntt Deterministic Games http://bit.ly/3GEMok9 CuuDuongThanCong.com Many possible formalizations, one is: t States: S (start at s0) t Players: P = {1 N} (usually take turns) t Actions: A (may depend on player / state) t Transition Function: S × A → S t Terminal Test: S → {t, f } t Terminal Utilities: S × P → R Solution for a player is a policy: S → A https://fb.com/tailieudientucntt Zero-Sum Games http://bit.ly/3GEMok9 Zero-Sum Games t Agents have opposite utilities (values on outcomes) t Lets us think of a single value that one maximizes and the other minimizes t Adversarial, pure competition CuuDuongThanCong.com General Games t Agents have independent utilities (values on outcomes) t Cooperation, indifference, competition, and more are all possible t More later on non-zero-sum games https://fb.com/tailieudientucntt Adversarial Search CuuDuongThanCong.com https://fb.com/tailieudientucntt Single-Agent Trees CuuDuongThanCong.com ··· ··· https://fb.com/tailieudientucntt Value of Game Tree Value of a State: utility of best achievable outcome from state Non terminal states: V (s) = maxs0 ∈kids(s) V (s0 ) CuuDuongThanCong.com ··· ··· https://fb.com/tailieudientucntt Terminal States: V (s) = known

Ngày đăng: 25/11/2022, 23:05