Model training method and device, resource allocation method, medium and electronic equipment
A technology for model training and resource allocation, applied in the computer field, can solve problems such as low resource utilization, inability to take into account local optimum and global optimum, unreasonable resource allocation, etc., and achieve the effect of improving resource utilization
Pending Publication Date: 2022-04-15
CEC CYBERSPACE GREAT WALL
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] To this end, the present application provides a model training method and device, resource allocation method, medium, and electronic equipment to solve the problem of inability to take into account both local optimality and global optimality when allocating resources, resulting in unreasonable resource allocation and resource utilization. low rate problem
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
[0042] The specific implementation manners of the present application will be described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementations described here are only used to illustrate and explain the present application, and are not intended to limit the present application.
[0043] The machine learning method is a method that uses existing data to generate a certain model and uses this model to make predictions. After the model structure and initial parameters are determined, the process of finding model parameters based on the model is called the model training process. During the training process of the model, it is necessary to iteratively adjust the parameters of the model according to the training data, so as to improve the prediction accuracy of the model. In practical applications, in order to obtain better prediction results, the training data set may be large, and the amount of data processing will...
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 discloses a model training method and device, a resource allocation method, a medium and electronic equipment, and the model training method comprises the steps: carrying out the iterative training of a preset resource allocation model at a set node, and obtaining a corresponding model parameter, an oracle value and an oracle verification value under the condition that each training reaches Nash equilibrium, the oracle verification value is determined based on the oracle value; determining whether Pareto efficiency is reached or not according to the oracle verification value of the set node and a preset value range; and when the Pareto efficiency is determined to be reached, stopping training the resource allocation model, and determining a target resource allocation model according to the model parameters obtained by the last training. In the model training process, global optimum and local optimum are considered through Nash equilibrium and Pareto efficiency, so that a more reasonable resource allocation scheme can be obtained when the resources are allocated based on the target resource allocation model, and the resource utilization rate is improved.
Description
technical field [0001] The present application relates to the field of computer technology, in particular to a model training method and device, resource allocation method, media, and electronic equipment. Background technique [0002] With the development of large-scale computing technology, in more and more cases, multiple nodes are required to jointly implement algorithms such as joint machine learning and federated learning. However, in the non-cooperative state, there is no interest agreement between nodes, and each node hopes to obtain the maximum benefit. Based on this, it is easy to cause multiple local optimums, but the overall learning result is not optimal. If algorithms such as joint machine learning and federated learning are applied to the field of resource allocation, the resources of some nodes will be optimal, but the global resource allocation will not be optimal, resulting in unreasonable resource allocation and low resource utilization. Therefore, how to...
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 Patent Type & Authority Applications(China)
IPC IPC(8): G06N20/00G06F9/50
CPCG06N20/00G06F9/5027
Inventor 张兴王峥瀛贾晓丰聂二保高嵩章敏朱江黎奇
Owner CEC CYBERSPACE GREAT WALL