Data Processing Integrating Singular Value Decomposition Algorithm and Tensor Chain Decomposition Algorithm

In the era of big data, processing information in multi-linear arrays is a challenge. In this paper, a parallel singular value decomposition algorithm based on unilateral Jacobi is proposed, and a data processing model of long and short memory network combined with the parallel tensor chain decomposition algorithm is constructed. The results show that the parallel efficiency of the algorithm reaches 0.95 with 30 cores, the compression ratio is 10, the accuracy and recall rate are 0.98 and 0.96, respectively. On the ImageNet dataset, model indicators are all over 0.9, showing excellent performance. The research not only improves the efficiency of data processing, but also provides new solutions for high-dimensional data analysis, especially in the aspects of feature extraction and dimensionality reduction. Combined with the advantages of LSTM in processing time series data, the overall performance of the model is improved.