Tensor Factorization
#Machine Learning #Factorization #Tensor
Tensors
We will be talking about tensors but we will skip the introduction to tensor for now.
In this article, we follow a commonly used convention for tensors in physics, the abstract index notation. We will denote tensors as $T^{ab\cdots}_ {\phantom{ab\cdots}cd\cdots}$, where the latin indices such as $^{a}$ are simply a placebo for the slot for this “tensor machine”. For a given basis (coordinate system), we can write down the components of this tensor $T^{\alpha\beta\cdots} _ {\phantom{\alpha\beta\cdots}\gamma\delta\cdots}$.
Published:
by L Ma;
L Ma (2019). 'Tensor Factorization', Datumorphism, 06 April. Available at: https://datumorphism.leima.is/wiki/machinelearning/factorization/tensorfactorization/.
Table of Contents
References:
 Anandkumar, A., Ge, R., Hsu, D., Kakade, S. M., & Telgarsky, M. (2012). Tensor decompositions for learning latent variable models. Journal of Machine Learning Research, 15(1), 2773–2832.
 Tensor Methods in Machine Learning
 Penrose graphical notation
 What is the practical difference between abstract index notation and “ordinary” index notation
 Tensor Decomposition: Fast CNN in your pocket
Current Ref:

wiki/machinelearning/factorization/tensorfactorization.md