Calculation graph evolution AI model automatic generation method based on structural similarity

A technology of structural similarity and automatic generation, applied in genetic models, calculations, computer components, etc., can solve problems such as low search efficiency of evolutionary algorithms, reduced sample diversity, poor model performance, etc., to improve model design efficiency, Increase the application range and enhance the effect of diversity

Pending Publication Date: 2020-02-07
BEIJING BENYING NETWORK TECH CO LTD
View PDF9 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In the process of network model search and design, there will be a large number of repeated network models with similar structures or similar performance. Because the calculation process of model performance is time-consuming, a large number of

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Calculation graph evolution AI model automatic generation method based on structural similarity

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0079] As described in step (1), the user prepares numerical calculation data, that is, platform production parameters (in csv format or image format), and the data contains label columns; set the maximum evolution generation of the model to 2 generations, and the number of model populations in each generation to 3; Presetting the configuration threshold for the structural similarity of the new model requires an edit distance of not less than 1; the smaller the preset fitness, the better the performance of the model; the fitness threshold of the calculated graph model——if the optimal model fitness is less than 40, it is considered to meet the end of evolution Condition, the calculation stops, when the evolutionary generation is greater than or equal to 2 generations, even if the fitness does not meet the above value, the evolution condition is considered to be met, and the calculation stops; the model with a preset fitness of more than 800 is considered an invalid model. The ca...

Embodiment 2

[0102] As described in step (1), the user prepares numerical calculation data, that is, platform production parameters (in csv format or image format), and the data contains label columns; define the generation number of each generation as G, and set the maximum evolution generation of the model to G max = 3 generations, the number of model populations in each generation is 5; the smaller the preset fitness, the better the performance of the model; the preset dynamic new model similarity threshold - the configuration threshold requires a similarity score not greater than G*0.1+0.5 (similarity The score is not greater than 0.8, that is, the similarity is not greater than 80%); calculate the fitness threshold of the graph model - if the optimal model fitness is less than 100, it is considered that the evolution end condition is met, and the calculation stops. When the evolution algebra is greater than or equal to 3 generations , even if the fitness does not meet the aforementione...

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

No PUM Login to view more

Abstract

The invention provides a calculation graph evolution AI model automatic generation method based on structural similarity. In the AI model automatic generation method based on calculation graph evolution, a graph similarity technology, namely a method of combining multiple graph similarity calculations, is adopted for calculating the similarity degree between a generated new network model and a known network model in the automatic generation process of an AI model, and a large number of repeated models, similar models and performance similar models are inhibited from appearing within a certainsimilarity threshold range. The diversity of model samples in the model search process can be effectively ensured, and the model network search success rate is improved. The performance degradation ofthe search network is obviously reduced. Meanwhile, the similarity threshold is dynamically adjusted according to the search efficiency, the function of jumping out of local optimum can be achieved according to model samples, and model search is accelerated. Therefore, the automatic generation efficiency of the AI model is improved.

Description

technical field [0001] The invention relates to the related technical field of AI models (AI models, namely artificial intelligence models), in particular to a method for automatically generating AI models based on structural similarity calculation graph evolution. Background technique [0002] Automatic generation of AI models is a cutting-edge research field. Automatic model generation can generate simpler and more efficient neural network models according to the distribution of data. The search space automatically generated by the AI ​​model is f n ×2 n(n-1) / 2 , where f is the number of different neuron operators, and n is the maximum depth of the neural network. It can be seen that during the generation process, with the increase of supported neural network operators and the deepening of the network model, the complexity of the problem may become a problem that approaches an infinite search space, resulting in the inability to solve it. [0003] At present, the main ...

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
no application Login to view more
IPC IPC(8): G06K9/62G06N3/04G06N3/08G06N3/12
CPCG06N3/086G06N3/126G06N3/045G06F18/22
Inventor 钱广锐宋煜傅志文吴开源
Owner BEIJING BENYING NETWORK TECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap