Skip to content

Latest commit

 

History

History
28 lines (19 loc) · 955 Bytes

README.md

File metadata and controls

28 lines (19 loc) · 955 Bytes

Forester Workspace

Forester is a rust crate for implementing various flavors of random forests and decision trees.

This workspace contains the forester library and examples in separate crates.

Please refer to the subdirectories for more detailed documentation.

Gallery

Classification

This example illustrates classification of three different labels.

The first image shows the decision regions created by a normal ExtraTreesClassifier. Artifacts resulting from axis-aligned splits are clearly visible.

The second image demonstrates how a random forest can utilize an infinite-dimensional feature space to produce the effect of roted splits which adapt much nicer to the shape of the data.

ExtraTreesRegressor Rotated Features

Regression

A simple example demonstrating non-linear regression with the ExtraTreesRegressor. Regression