AI model automatic generation method based on computational graph evolution

A technology for automatically generating and calculating graphs, applied in computing models, genetic models, calculations, etc., can solve problems such as smaller network differences and long model search processes, so as to prevent performance decline, ensure efficiency, and improve model design efficiency Effect

Active Publication Date: 2019-05-21
BEIJING BENYING NETWORK TECH CO LTD
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In many cases, when the search is approaching the optimal solution, the network difference is also becoming smaller, and th

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
  • AI model automatic generation method based on computational graph evolution

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0056] As described in step (1), the user prepares numerical calculation data (in csv format or image format), and the data contains label columns; set the maximum evolution generation of the model to 3 generations, and the number of model populations in each generation is 5; preset fitness The smaller the model, the better the performance; calculate the fitness threshold of the graph model - if the optimal model fitness is less than 50, it is considered to meet the evolution end condition, and the calculation stops; the model with a preset fitness of more than 1000 is considered an invalid model.

[0057] As described in step (2), use the genetic random operator to randomly generate 5 models of the first generation, respectively: randomly generate 5 models of the first generation, respectively: computing graph model 1, computing graph model 2, computing graph model 3. Computing graph model 4 and computing graph model 5.

[0058] As described in step (3), encode the model expr...

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 an AI model automatic generation method based on computational graph evolution. The method mainly comprises the following steps: presetting data; Utilizing a genetic algorithm operator to generate a first-generation calculation graph model and calculating model performance according to a calculation graph structure; removing the invalid model and the repeated model, and reserving the remaining models as alternative models as seeds of the next generation; Selecting a plurality of optimal models; The alternative model generates a calculation graph new model by using a genetic algorithm operator; judging whether the new calculation graph model generated in the last step is generated or not; storing the new model as a new generation of computational graph model, and judging whether the new model meets preset data and evolutionary ending conditions or not; and summarizing evolutionary calculation results, and selecting an optimal model. According to the invention, machine learning and deep learning can be carried out simultaneously. The repeated calculation frequency of the same model is avoided, and the model design efficiency is improved; Jumping out of local optimum; Declining of the performance of the search network is prevented. The evaluation can be directly carried out without training through actual data.

Description

technical field [0001] The present invention relates to the technical field related to AI models (AI models are artificial intelligence models), in particular to a method for automatically generating AI models based on computational 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 networks based on 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, the complexity of the problem may become a problem approaching an infinite search space, which makes it impossible to solve. [0003] At present, the main search methods include reinforcement lea...

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): G06N20/00G06N3/12
CPCG06N3/12G06N20/00
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