Tensor Decomposition
[Academic Review] An exploration of high-dimensional data decompositions and their practical applications in machine learning.
This presentation evaluates the mathematical foundations of tensor decomposition, specifically focusing on CP (CANDECOMP/PARAFAC) and Tucker decompositions. The material includes a detailed breakdown of the underlying linear algebra and least squares optimization problems. Furthermore, it presents practical sample cases demonstrating the implementation and utility of each approach in machine learning workflows.
Click here to view the embedded file.