Tensor decomposition: A versatile method for heterogeneous biological data fusion
Koki Tsuyuzaki · 2020
Data in life sciences are diverse, and it is challenging to know how to represent these heterogeneous datasets, what algorithms to apply to relate them to each other, and how to make biological interpretations of such datasets. In this talk, we will introduce tensor decomposition as a potential unified framework to solve these problems. Next, we will introduce scTensor package, which is an application of tensor decomposition to single-cell RNA-seq datasets. Finally, we will introduce DelayedTensor package, which is an out-of-core and sparse implementation of tensor format to deal with extremely large-scaled tensor data.