In my last post I talked about two types of machine learning algorithms – supervised and unsupervised. The linear regression model comes under the first category. 738 more words
Tags » Machine Learning
The Toy Model
First consider the following toy problem:
- the input data distribution is a mixture of very many (perhaps infinite) components of localized distributions (e.g. 490 more words
In the Stanford Deep Learning tutorial, whitening is introduced as a powerful pre-processing step for feature learning.
There is an interesting paper published by Adam Coates and Andrew Ng from Stanford and Honglak Lee from the University of Michigan where, among other findings, they demonstrate the important role of whitening for feature learning. 1,434 more words
I ran across an unusual scenario recently where it was beneficial to “de-modularize” some code. I’m not sure if “de-modularize” is a word or not, but I mean refactoring some code that was in two functions to one larger function. 610 more words