NeuralNetwork
Class representing a Neural Network
Constructor Summary
| Public Constructor | ||
| public |
constructor(inputNodes: Int, hiddenNodes: Int, outputNodes: Int, learningRate: Int) Create a NeuralNetwork |
|
Method Summary
| Public Methods | ||
| public |
Returns a copy of the NeuralNetwork |
|
| public |
Evaluate inputs based on the current state of the NeuralNetwork |
|
| public |
Train the NeuralNetwork |
|
Public Constructors
public constructor(inputNodes: Int, hiddenNodes: Int, outputNodes: Int, learningRate: Int) source
Create a NeuralNetwork
Params:
| Name | Type | Attribute | Description |
| inputNodes | Int | The number of input nodes |
|
| hiddenNodes | Int | The number of hidden nodes |
|
| outputNodes | Int | The number of output nodes |
|
| learningRate | Int | The learning rate for training |
Public Methods
public eval(inputArr: array[number]): object source
Evaluate inputs based on the current state of the NeuralNetwork
Params:
| Name | Type | Attribute | Description |
| inputArr | array[number] | A single dimensional array of number inputs |
Return:
| object | An object containing matricies representing the input, hidden node outputs, and final outputs |
public train(inputArr: array[number], targetArr: array[number]): object source
Train the NeuralNetwork
Params:
| Name | Type | Attribute | Description |
| inputArr | array[number] | A single dimensional array of number inputs |
|
| targetArr | array[number] | A single dimensional array of numbers representing expected outputs |
Return:
| object | An object containing properties representing errors between the Network's output and target outputs and errors in the hidden node layer |