Machine Learning

Day 97: Locally weighted regression

Locally weighted regression is a very powerful non-parametric model used in statistical learning.

Hopfield net is a kind of neural network with binary threshold units and recurrent connections. The connections are symmetric and the model has a global energy function which results in a surprisingly simple training algorithm.

Logistic regression is a regression model used for data classification. The model is very simple since, in fact, it is a linear classifier.

However, when you start to explore your data, the simplest model is a good spot to start at.