Artificial intelligence AI model training method and device, equipment and medium

Abstract

The invention discloses an artificial intelligence AI model training method and device, equipment and a medium, and relates to the field of artificial intelligence machine learning. The method comprises the steps that an artificial intelligence AI model is called to conduct game match in a game program to obtain training data, wherein the training data comprises a reference game state in the gamematch, a target game action output by a decision network according to the reference game state and a state value output by a value network according to the reference game state, the state value comprises k state sub-values on k value classifications, and k is an integer greater than 1; according to the training data and k value calculation formulas corresponding to the k value classifications, theaction value of the target game action adopted by the artificial intelligence AI model in the reference game state is calculated, wherein the action value comprises k action sub-values on the k valueclassifications; and the artificial intelligence AI is trained model according to the difference between the state value and the action value. The method can improve the accuracy of estimating the state value of the value network.

Description

technical field [0001] The embodiments of the present application relate to the field of machine learning of artificial intelligence, and in particular to a training method, device, equipment and medium of an artificial intelligence AI model. Background technique [0002] Reinforcement learning is one of the paradigms and methodologies of machine learning, which is used to describe and solve the problem that an agent (agent, also known as "agent") learns strategies to maximize rewards or achieve specific goals in the process of interacting with the environment. The AI ​​(Artificial Intelligence, artificial intelligence) model designed based on reinforcement learning can make game decisions to win the game. [0003] The AI ​​model includes a decision network and a value network. The decision network is used to determine action instructions based on the game state, and the value network is used to evaluate the value of the game state. In the training phase of the AI ​​model, i...

AI Technical Summary

Problems solved by technology

[0005] However, due to the complexity of the game environment of MOBA games and the change of game parameters due to the execution of actions, the changes of different game parameters have different influences on strategic decisions. The impact length of the tower being destroyed

The method in the related art only considers the time decay, the influence length of the value factors of all game parameters is the same, and the value network cannot accurately state the value

Images