Tensor projection mechanism and algorithm implementation

• Published in: Applied intelligence (Dordrecht, Netherlands) Vol. 51; no. 11; pp. 8176 - 8191
• Main Authors: Wang, Fei, Wang, Zhanchang, Li, Xiang, Chen, Yu
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.11.2021; Springer Nature B.V
• Subjects: Algorithms; Artificial Intelligence; BeiDou Navigation Satellite System; Communication satellites; Computer Science; Machine learning; Machines; Manufacturing; Mathematical analysis; Mechanical Engineering; Parallel processing; Processes; Projection; Quantum computing; Source code; Tensors; Transformations; Parallelization; Quantum computation language; Tensor projection; Numerical algorithm
• ISSN: 0924-669X; 1573-7497
• DOI: 10.1007/s10489-021-02332-3

This paper analyses the mechanism of tensor projection transformation in depth and introduces a high-efficiency original algorithm developed in a quantum computing language for forward and backward projection between multidimensional tensors and one-dimensional vectors. Additionally, the author compares this algorithm with similar methods from both the Python scientific computing package and other relative development kits in method calls and source code to demonstrate the innovation of the tensor projection algorithm. On this basis, the classical convolution operation program commonly used in machine learning has been parallelized and improved, the analysis algorithm of the Beidou communication satellite view area has been parallelized and improved, and the actual operating efficiency has been greatly improved. After verification, the tensor projection transformation successfully solves the problem of location index mapping of the entity units among different dimensions, can provide a means for optimizing the traditional model of traversal algorithm, and can have significant reference value in eigenspace transformation against a tensor field.