When Machine Learning Meets AI and Game Theory
نویسندگان
چکیده
We study the problem of development of intelligent machine learning applications to exploit the problems of adaptation that arise in multi-agent systems, for expected-long-termprofit maximization. We present two results. First, we propose a learning algorithm for the Iterated Prisoners Dilemma (IPD) problem. Using numerical analysis we show that it performs strictly better than the tit-for-tat algorithm and many other adaptive and non-adaptive strategies. Second, we study the same problem from the aspect of zero-sum games. We discuss how AI and Machine Learning techniques work closely to give our agent a ’mind-reading’ capability.
منابع مشابه
On-Line Adaptation of Game Opponent AI in Simulation and in Practice
Unsupervised online learning in commercial computer games allows computer-controlled opponents to adapt to the way the game is being played, thereby providing a mechanism to deal with weaknesses in the game AI and to respond to changes in human player tactics. For online learning to work in practice, it must be fast, effective, robust, and efficient. This paper proposes a novel technique called...
متن کاملEducational Advances in Artificial Intelligence
WINTER 2013 127 (search, game theory, logic, robotics, and others) could be used to serve as an introduction to CS and computation. During the EAAI teaching and mentoring workshop, participants were introduced to the basic theory of active learning. Example active learning activities were described, then participants worked together to construct additional examples of methods that could be inco...
متن کاملOnline Adaptation of Computer Game Opponent AI
Online learning in commercial computer games allows computer-controlled opponents to adapt to human player tactics. For online learning to work in practice, it must be fast, effective, robust, and efficient. This paper proposes a technique called “dynamic scripting” that meets these requirements. In dynamic scripting an adaptive rule-base is used for the generation of intelligent opponents on t...
متن کاملKnowledge acquisition for adaptive game AI
Game artificial intelligence (AI) controls the decision-making process of computercontrolled opponents in computer games. Adaptive game AI (i.e., game AI that can automatically adapt the behaviour of the computer players to changes in the environment) can increase the entertainment value of computer games. Successful adaptive game AI is invariably based on the game’s domain knowledge. We show t...
متن کاملMechanism Learning with Mechanism Induced Data
Machine learning and game theory are two important directions of AI. The former usually assumes data is independent of the models to be learned; the latter usually assumes agents are fully rational. In many modern Internet applications, like sponsored search and crowdsourcing, the two basic assumptions are violated and new challenges are posed to both machine learning and game theory. To better...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2012