Non-parametric regression

This package provides a set of non-parametric methods for regression. The algorithmic implementation of the methods are in C++ whilst the interface is in Python. Non-parametric regression methods typically retains all the training data. The value of a new test point is a function of the neighboring points in the input space. As a result it is essential to use a fast space partition method, such as FLANN. Current implemented method are:

  • Locally Weighted Regression (LWR)