DEEP NEURAL NETWORK INTERPRETATION METHOD AND DEVICE, TERMINAL, AND STORAGE MEDIUM

摘要

A deep neural network interpretation method and device, a terminal, and a storage medium, relating to the technical field of artificial intelligence. Specifically disclosed are: determining all pieces of target data among multiple input pieces of data according to a result output by a deep neural network, inputting the target data into an interpretation factor prediction model, separately to obtain a group of interpretation factors corresponding to each piece of target data, and then counting and sorting the number of occurrences of each interpretation factor in multiple groups of interpretation factors, and selecting a target number of highest-ranking interpretation factors as target interpretation factors of a preset target category. According to the method, an interpretation factor of a prediction result is obtained by means of a pre-trained interpretation factor prediction model, so as to interpret a dominant factor affecting the prediction result of the deep neural network. The method can be applied to a scene of intelligent government affairs/smart city management/intelligent community/intelligent security/intelligent logistics/intelligent healthcare/intelligent education/intelligent environmental protection/intelligent transport, so as to promote the construction of smart city.