Video game decision-making method based on auxiliary task learning
A technology for auxiliary tasks and video games, applied in video games, neural learning methods, indoor games, etc., can solve problems such as incompleteness
Active Publication Date: 2020-06-09
HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)
View PDF8 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
At this time, in order to obtain useful reward values for the agent to learn long-term actions from the complex 3D video game signal input, it is not comprehensive enough to consider the problem only from the perspective of using intrinsic rewards to increase the source of reward information
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0026] The present invention will be further described below in conjunction with the description of the drawings and specific embodiments.
[0027] It is the core content of the present invention to apply the deep reinforcement learning method, combined with the advanced internal reward mechanism, to form a decision-making model and technology with a certain level of intelligence, so that the game agent can obtain high scores in video games.
[0028] Such as figure 1 As shown, a video game decision-making method based on auxiliary task learning includes the following steps:
[0029] S1. Construct a neural grid model;
[0030] S2. Starting a multi-process video game environment;
[0031] S3, judging whether the specified round has been run, if not, then enter step S4, if yes, then enter step S6;
[0032] S4. Obtain game experience and update the experience pool;
[0033] S5. Input the experience into the neural grid model, update the parameters of the neural grid model, and...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention provides a video game decision-making method based on auxiliary task learning. The video game decision-making method comprises the following steps: S1, constructing a neural network model; S2, starting a multi-process video game environment; S3, judging whether a specified round is operated or not, if not, entering the step S4, and if so, entering the step S6; S4, obtaining game experience, and updating the experience pool; S5, inputting experience into the neural grid model, updating parameters of the neural grid model, and returning to the step S3; S6, storing the neural network model; S7, making a decision in the video game by using the neural grid model; S8, ending. The method has the beneficial effect that the state value in the three-dimensional scene and the action ofthe intelligent agent causing the state change can be estimated more accurately.
Description
technical field [0001] The invention relates to a video game decision-making method, in particular to a video game decision-making method based on auxiliary task learning. Background technique [0002] Video games appeared in the early 1970s. Since the birth of video games, the technology of automatic decision-making of intelligent bodies in video games through artificial intelligence technology has always been a research hotspot in industry and academia, and has huge commercial value. . In recent years, the rapid development of deep reinforcement learning methods has provided an effective way to realize this technology. Generally speaking, game decision-making techniques are only as good as how many points are scored in the game or whether the game can be won, and the same is true for video games. [0003] The development of artificial intelligence technology is changing with each passing day, and machine game, as one of the hot research fields, has received extensive att...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06N3/04G06N3/08A63F13/45A63F13/46
CPCG06N3/049G06N3/08A63F13/45A63F13/46G06N3/045
Inventor 王轩张加佳漆舒汉曹睿杜明欣刘洋蒋琳廖清夏文李化乐
Owner HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN (INSTITUTE OF SCIENCE AND TECHNOLOGY INNOVATION HARBIN INSTITUTE OF TECHNOLOGY SHENZHEN)