Research on Task Correlation of Artificial Intelligence Datasets Based on Stacked Sparse Auto Encoder Networks

In order to solve the problem of high time and space complexity of task correlation computation in datasets, this contribution proposes a method of task correlation computation in artificial intelligence datasets based on Stack Sparse Auto Encoder Networks (SSAE). This method extracts highly abstract low-dimensional features through the powerful feature extraction and representation ability of SSAE, and provides a reliable basis for the quality assessment of data sets by using the European distance quantization task correlation evaluation index between the distribution of abstract low-dimensional features and the benchmark data sets. Experimental results show that the proposed feature extraction method can effectively extract features in layers, and the distribution of abstract lowdimensional features in layers can be improved. Experimental results show that the proposed feature extraction method can effectively extract features in layers, which can improve the accuracy of classification and recognition while reducing the computational complexity of task correlation..