Momentum

java.lang.Object
- me.yixqiao.jlearn.optimizers.Optimizer
- - me.yixqiao.jlearn.optimizers.Momentum

```
public class Momentum
extends Optimizer
```
Gradient descent with momentum.

Constructor Summary

Constructors
Constructor Description

Momentum(double learningRate, double momentum)
Create a new instance.

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method	Description
`Matrix`	`apply(Matrix g)`	Apply the optimizer.
`Optimizer`	`cloneSettings()`	Return a clone of this optimizer's settings.
`void`	`multiplyLR(double d)`	Multiply the learning rate.

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- Constructor Detail
  - Momentum
```
public Momentum(double learningRate,
                double momentum)
```
    Create a new instance.
    
    Parameters:
    
    learningRate - learning rate
    
    momentum - momentum (0-1, higher value means more momentum)
- Method Detail
  - cloneSettings
```
public Optimizer cloneSettings()
```
    Description copied from class: Optimizer
    
    Return a clone of this optimizer's settings.
    
    Specified by:
    
    cloneSettings in class Optimizer
    
    Returns:
    
    the clone
  - multiplyLR
```
public void multiplyLR(double d)
```
    Description copied from class: Optimizer
    
    Multiply the learning rate.
    Used to scale the learning rate in different batch sizes.
    
    Specified by:
    
    multiplyLR in class Optimizer
    
    Parameters:
    
    d - amount
  - apply
```
public Matrix apply(Matrix g)
```
    Description copied from class: Optimizer
    
    Apply the optimizer.
    
    Specified by:
    
    apply in class Optimizer
    
    Parameters:
    
    g - errors calculated from backpropagation.
    
    Returns:
    
    the gradients from the optimizer