blog/matrix-inverse-mini-batch/

[utterances-bot]

Matrix inverse mini-batch updates • Max Halford

The inverse covariance matrix, also called precision matrix, is useful in many places across the field of statistics. For instance, in machine learning, it is used for Bayesian regression and mixture modelling. What’s interesting is that any batch model which uses a precision matrix can be turned into an online model. That is, provided the precision matrix can be estimated in a streaming fashion. For instance, scikit-learn’s elliptic envelope method could have an online variant with a partial_fit method.

https://maxhalford.github.io/blog/matrix-inverse-mini-batch/

[smastelini]

Nice article, Max! This sounds like the future! Gaussian Process also comes to my mind :)