Abstract
The proposed model represents an original approach to general game playing, and aims at creating a player able to develop a strategy using as few requirements as possible, in order to achieve the maximum generality. The main idea is to modify the known minimax search algorithm removing its task-specific component, namely the heuristic function: this is replaced by a neural network trained to evaluate the game states using results from previous simulated matches. A method for simulating matches and extracting training examples from them is also proposed, completing the automatic procedure for the setup and improvement of the model. Part of the algorithm for extracting training examples is the Backward Iterative Deepening Search, a new original search algorithm which aims at finding, in a limited time, a high number of leaves along with their common ancestors.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
In the case of the proposed system, since the output is real-valued, the equality constraint must be softened and all values within a margin \(\varepsilon \) are accepted.
- 2.
The complete tree of a game is the tree having as root the initial state and as children of a node all the states reachable with a single legal move from that node. The leaves of this tree are the terminal states of the game.
- 3.
The most common commercial versions of this game are played on a suspended grid, where tokens fall until the lowest free position.
- 4.
The branching factor is the average number of branches (successors) from a (typical) node in a tree. It indicates the bushiness and hence the complexity of a tree. If a tree branching factor is B, then at depth d there will be approximately \(B^d\) nodes.
References
Silver, D., Huang, A., et al.: Mastering the game of Go with deep neural networks and tree search. Nature 529(7587), 484–489 (2016)
Draper, S., Rose, A.: Sancho GGP player. http://sanchoggp.blogspot.com
Michulke, D.: Neural networks for high-resolution state evaluation in general game playing. In: IJCAI-11 Workshop on General Game Playing (GIGA11), pp. 31–37 (2011)
Schiffel, S., Thielscher, M.: Fluxplayer: a successful general game player. In: 22nd National Conference on Artificial intelligence, pp. 1191–1196. AAAI Press, Menlo Park (2007)
Świechowski, M., Mańdziuk, J.: Specialized vs. multi-game approaches to AI in games. In: Angelov, P., et al. (eds.) Intelligent Systems 2014. AISC, vol. 322, pp. 243–254. Springer, Heidelberg (2015). doi:10.1007/978-3-319-11313-5_23
Świechowski, M., Park, H., Mańdziuk, J., Kim, K.-J.: Recent advances in general game playing. Sci. World J. 2015, Article ID 986262, 22 p. (2015). doi:10.1155/2015/986262
Schmidt, W.F., Kraaijveld, M., Duin, R.P.W., et al.: Feedforward neural networks with random weights. In: International Conference on Pattern Recognition, Conference B: Pattern Recognition Methodology and Systems, pp. 1–4 (1992)
Pao, Y.H., Park, G.H., Sobajic, D.J.: Learning and generalization characteristics of the random vector functional-link net. Neurocomputing 6, 163–180 (1994)
Huang, G.B., Chen, L., Siew, C.K.: Universal approximation using incremental constructive feedforward networks with random hidden nodes. IEEE Trans. Neural Netw. 17, 879–892 (2006)
Liang, N.Y., Huang, G.B., Saratchandran, P., Sundararajan, N.: A fast and accurate online sequential learning algorithm for feedforward networks. IEEE Trans. Neural Netw. 17, 1411–1423 (2006)
Russell, S., Norvig, P.: Artificial Intelligence: A Modern Approach. Prentice-Hall, Egnlewood Cliffs (1995)
Penrose, R.: On best approximate solutions of linear matrix equations. Math. Proc. Camb. Philos. Soc. 52, 17–19 (1956)
Bishop, C.: Pattern Recognition and Machine Learning. Springer, Heidelberg (2006)
Gherrity, M.: A game-learning machine. Ph.D. thesis, University of California, San Diego (1993)
Allis, L.W.: A knowledge-based approach of connect-four. Technical report, Vrije Universiteit, Subfaculteit Wiskunde en Informatica (1988)
British Othello Federation: Game Rules. http://www.britishothello.org.uk/rules.html
Cirasella, J., Kopec, D.: The History of Computer Games. CUNY Academic Works, New York (2006)
Mitchell, D.H.: Using features to evaluate positions in experts’ and novices’ Othello games. Masters thesis, Northwestern University, Evanston (1984)
MacGuire, S.: Strategy Guide for Reversi and Reversed Reversi. www.samsoft.org.uk/reversi/strategy.htm#position
Mnih, V., Kavukcuoglu, K., Silver, D., et al.: Human-level control through deep reinforcement learning. Nature 518(7540), 529–533 (2015)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing AG
About this paper
Cite this paper
Ghignone, L., Cancelliere, R. (2016). Neural Learning of Heuristic Functions for General Game Playing. In: Pardalos, P., Conca, P., Giuffrida, G., Nicosia, G. (eds) Machine Learning, Optimization, and Big Data. MOD 2016. Lecture Notes in Computer Science(), vol 10122. Springer, Cham. https://doi.org/10.1007/978-3-319-51469-7_7
Download citation
DOI: https://doi.org/10.1007/978-3-319-51469-7_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-51468-0
Online ISBN: 978-3-319-51469-7
eBook Packages: Computer ScienceComputer Science (R0)