ICA.ON3Trainer - Java-Scheme Laboratory

Skip navigation links

All Classes

SEARCH:

Summary:
Nested |
Field |
Constr |
Method

Detail:
Field |
Constr |
Method

java.lang.Object
- samer.tools.AnonymousTask
- - samer.models.ICA.ON3Trainer

All Implemented Interfaces:

Model.Trainer, Task

Direct Known Subclasses:

ICA.NewtonTrainer, ICAWithScaler.DifferentialTrainer

Enclosing class:

ICA
```
public class ICA.ON3Trainer
extends AnonymousTask
implements Model.Trainer
```
This trainer saves on an O(N^2) step during accumulation, at the expense of an O(N^3) flush. As long as the batch size is O(N), then it should be about the same overall. The advantage is the collected statistics are more transparent, and can be used to make scalar or diagonal updates more frequenty.

See Also:

Serialized Form

Constructor Summary

Constructors
Constructor and Description

ON3Trainer()

Constructors
Constructor and Description
`ON3Trainer()`

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`void`	`accumulate()` collect statistics for parameter update
`void`	`accumulate(double w)` weighted accumulate
`void`	`dispose()`
`void`	`flush()` use collected stats to update parameters and reset
`Matrix`	`getGMatrix()` this is so you can manipulate the matrix before flushing
`void`	`oneshot()` Must be equivalent to reset(); accumulate(); flush(); but can be optimised for non-batched training
`void`	`reset()` reset accumulators without updating parameters
`void`	`run()`
`void`	`starting()`
`java.lang.String`	`toString()`

Methods inherited from class samer.tools.AnonymousTask
stopping

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

- Constructor Detail
  - ON3Trainer
```
public ON3Trainer()
```
- Method Detail
  - toString
```
public java.lang.String toString()
```
    Overrides:
    
    toString in class java.lang.Object
  - getGMatrix
```
public Matrix getGMatrix()
```
    this is so you can manipulate the matrix before flushing
  - starting
```
public void starting()
```
    Specified by:
    
    starting in interface Task
    
    Overrides:
    
    starting in class AnonymousTask
  - run
```
public void run()
```
    Specified by:
    
    run in interface Task
  - dispose
```
public void dispose()
```
    Specified by:
    
    dispose in interface Model.Trainer
    
    Specified by:
    
    dispose in interface Task
    
    Overrides:
    
    dispose in class AnonymousTask
  - oneshot
```
public void oneshot()
```
    Description copied from interface: Model.Trainer
    
    Must be equivalent to reset(); accumulate(); flush(); but can be optimised for non-batched training
    
    Specified by:
    
    oneshot in interface Model.Trainer
  - reset
```
public void reset()
```
    Description copied from interface: Model.Trainer
    
    reset accumulators without updating parameters
    
    Specified by:
    
    reset in interface Model.Trainer
  - accumulate
```
public void accumulate()
```
    Description copied from interface: Model.Trainer
    
    collect statistics for parameter update
    
    Specified by:
    
    accumulate in interface Model.Trainer
  - accumulate
```
public void accumulate(double w)
```
    Description copied from interface: Model.Trainer
    
    weighted accumulate
    
    Specified by:
    
    accumulate in interface Model.Trainer
  - flush
```
public void flush()
```
    Description copied from interface: Model.Trainer
    
    use collected stats to update parameters and reset
    
    Specified by:
    
    flush in interface Model.Trainer

Skip navigation links

All Classes

Summary:
Nested |
Field |
Constr |
Method

Detail:
Field |
Constr |
Method

Java-Scheme Laboratory

Class ICA.ON3Trainer

Constructor Summary

Method Summary

Methods inherited from class samer.tools.AnonymousTask

Methods inherited from class java.lang.Object

Constructor Detail

ON3Trainer

Method Detail

toString

getGMatrix

starting

run

dispose

oneshot

reset

accumulate

accumulate

flush