Module rusty_machine::learning::lin_reg

Linear Regression module

Contains implemention of linear regression using OLS and gradient descent optimization.

The regressor will automatically add the intercept term so you do not need to format the input matrices yourself.

Usage

use rusty_machine::learning::lin_reg::LinRegressor;
use rusty_machine::learning::SupModel;
use rusty_machine::linalg::Matrix;
use rusty_machine::linalg::Vector;

let inputs = Matrix::new(4,1,vec![1.0,3.0,5.0,7.0]);
let targets = Vector::new(vec![1.,5.,9.,13.]);

let mut lin_mod = LinRegressor::default();

// Train the model
lin_mod.train(&inputs, &targets).unwrap();

// Now we'll predict a new point
let new_point = Matrix::new(1,1,vec![10.]);
let output = lin_mod.predict(&new_point).unwrap();

// Hopefully we classified our new point correctly!
assert!(output[0] > 17f64, "Our regressor isn't very good!");

Structs

LinRegressor

Linear Regression Model.