Machine learning algorithm based on tensor networks is a new direction that has just developed in recent years. This direction could help us better understand the complexity of traditional machine learning datasets, the expressive power of traditional machine learning models, such as probability graph models, and it could also be regarded as a classical simulation for a large class of quantum machine learning models. In this talk, I will first give a brief overview of both tensor networks and traditional machine learning. Then, starting from the similarity of data and models, I will explain why tensor networks models originally used for quantum many-body research can be connected with traditional machine learning. After that, I will specifically introduce several typical tensor network machine learning models that have been proposed. Finally, I will summarize the significance of this direction and the future development.

Online ID:TBA
Password: TBA