Business execution method and device and related equipment thereof

Abstract

The invention discloses a service execution method, device and system and a computer readable storage medium. According to the service execution method, an existing supervised learning game model is utilized to perform self-game; the game model is corrected according to the game result in the game process; and according to the corrected game model, a corresponding game sample for continuous training of a subsequent supervised learning game model is generated, so that the game level of the supervised learning game model is gradually improved by optimizing the game sample, the model precision isensured, and the accuracy of a game business direct result is further improved.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, in particular to a business execution method, and also to a business execution device, system, and computer-readable storage medium. Background technique [0002] Machine games, known as the fruit fly of artificial intelligence, have been at the forefront of artificial intelligence research. For example, the poker game is a typical game of incomplete information. It is a long-standing challenge in artificial intelligence research. By using supervised learning to replicate human player decisions , many game intelligence systems have reached the level of advanced players. However, although an end-to-end game strategy model can be obtained by using neural network for supervised learning based on human game data, the performance level of the game strategy model trained only based on human game data is limited by the quality of training data. Because the human player samples...

AI Technical Summary

Problems solved by technology

However, although the end-to-end game strategy model can be obtained by using neural network for supervised learning based on human game data, the performance level of the game strategy model trained only based on human game data is limited by the quality of the training data. Because human player samples contain strategic error data, the quality of the sample data set limits the further improvement of the performance of the learned network model to a certain extent, resulting in low accuracy of the model, thereby reducing the accuracy of the execution results of the corresponding game business

Images