Trait vikos::Model

pub trait Model: Clone {
    type Input;
    type Target: Float;
    fn predict(&self, &Self::Input) -> Self::Target;
    fn num_coefficents(&self) -> usize;
    fn gradient(&self, coefficent: usize, input: &Self::Input) -> Self::Target;
    fn coefficent(&mut self, coefficent: usize) -> &mut Self::Target;
}

A Model is defines how to predict a target from an input

A model usually depends on several coefficents whose values are derived using a training algorithm

Associated Types

type Input

Input features

type Target: Float

Target type

Required Methods

fn predict(&self, &Self::Input) -> Self::Target

Predicts a target for the inputs based on the internal coefficents

fn num_coefficents(&self) -> usize

The number of internal coefficents this model depends on

fn gradient(&self, coefficent: usize, input: &Self::Input) -> Self::Target

Value predict derived by the n-th coefficent at input

fn coefficent(&mut self, coefficent: usize) -> &mut Self::Target

Mutable reference to the n-th coefficent

Implementors

• impl<I> Model for Constant<I>
• impl<V: Vector> Model for Linear<V> where V::Scalar: Float
• impl<V: Vector> Model for Logistic<V> where V::Scalar: Float