Program detection method and device based on neural network attention mechanism and medium

The invention relates to a program detection method and device based on a neural network attention mechanism and a medium. The method comprises the following steps: step 1) obtaining an initialized seed case; 2) training the seed case by using a feedforward neural network model combined with an attention mechanism, and guiding mutation by using an AdaMax gradient descent algorithm according to the trained feedforward neural network model to obtain a test case; 3) inputting the test case into the to-be-tested target program, and obtaining a test result based on a fuzzy test method; 4) performing iterative optimization on the feedforward neural network model according to a test result of the step 3) to obtain an optimal test case; and 5) inputting the optimal test case into the target program to be tested, and obtaining a final test result based on a fuzzy test method. Compared with the prior art, the method has the advantages of being high in detection guidance quality, high in detection automation degree, lar