Abstract state model construction method based on recurrent neural network

A technology of cyclic neural network and state model, which is applied in the direction of biological neural network model, neural learning method, neural architecture, etc., can solve problems such as difficult testing of the internal state of cyclic neural network, and achieve the effect of solving difficult testing problems and simplifying analysis

Pending Publication Date: 2020-10-30
深圳慕智科技有限公司
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] In the deep learning test, there are loops inside the recurrent neural network, which makes it difficult to test the internal state

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
  • Abstract state model construction method based on recurrent neural network
  • Abstract state model construction method based on recurrent neural network
  • Abstract state model construction method based on recurrent neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0016] This patent implements the test of the cyclic neural network model through the abstract state model construction technology. The abstract model construction technology is mainly used. The specific key technologies involved are: cyclic neural network (RNN), state vector acquisition technology, state vector discretization technology, State transition probability calculation technology, abstract state model construction technology, etc.

[0017] 1. State vector acquisition

[0018] In the present invention, we use the values ​​of the internal neurons of the cyclic neural network to arrange in order to form the state vector of the cyclic neural network. We give the internal state and state relationship of the recurrent neural network model in a formal way, and use this as a basis to describe its characteristics. A neural network can be abstractly represented as a differentiable parameterized function f() whose input can be vectorized as x∈X,. After the cyclic neural netwo...

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

An abstract state model construction method based on a recurrent neural network is characterized in that after a recurrent neural network model, namely test data, is provided, abstraction is carried out according to the internal state and the transfer rule of the recurrent neural network model to generate a corresponding abstract state model. The method comprises three parts: state vector acquisition, state vector discretization and abstract state model construction. An internal state vector of the recurrent neural network is simplified and analyzed in a discretization mode, and a state modelis abstracted according to an internal state and a transfer rule of the recurrent neural network. Finally, an abstract state transition model is obtained, and the model reflects abstract state characteristics of the original recurrent neural network.

Description

technical field [0001] The invention belongs to the field of software testing, in particular to a testing method of a cyclic neural network. After the cyclic neural network model and test data are provided, the corresponding abstract state model is generated by abstracting according to its internal state and transition rules. Background technique [0002] Deep learning has made significant progress in many practical fields, such as image processing, speech recognition, natural language processing, and autonomous driving. However, state-of-the-art deep learning systems still suffer from quality, reliability, and safety issues, which can lead to accidents and catastrophic events. Early testing of deep learning systems is of great significance for discovering defects and improving system quality. Although the analysis process and testing techniques of traditional software are mature, the existing tools cannot be directly applied to deep learning systems. In order to check th...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06F11/36G06N3/04G06N3/08
CPCG06F11/3672G06N3/049G06N3/08G06N3/047G06N3/044G06N3/045
Inventor 陈振宇高新宇刘佳玮尹伊宁
Owner 深圳慕智科技有限公司
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