Test data generation method of abstract state model based on recurrent neural network

Abstract

A test data generation method of an abstract state model based on a recurrent neural network is characterized in that after the 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, a fuzzy test is carried out on the generated abstract model, and therefore potential defects are found. The method comprises four components: an abstract model construction module, a fuzzy test module, a coverage standard definition moduleand a data display module. The recurrent neural network model and the sample data are uploaded, and a system can start fuzzy test generation after giving a neural network and the sample data. The system performs model abstraction on the recurrent neural network to construct an abstract state model; the original sample data is mutated through a mutation algorithm, so that a large amount of test data is generated. The test is completed under a specific coverage standard.

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 recurrent neural network model and test data are provided, the corresponding abstract state model is abstracted according to its internal state and transition law, and the generated abstract model is fuzzy tested to find potential defects. 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 d...

AI Technical Summary

Problems solved by technology

[0007] The problem to be solved by the present invention is: in the deep learning test, there is a loop inside the cyclic neural network, which makes the internal state difficult to test.

Images