Trying to setup a Neural Network, having some trouble with structure

Question

I am trying to setup a very simple neural network, which will use a multilayered perceptron approach.

I have 4 classes:

Node: which only has a constructor with nothing in it, and an output method (which I still have to define)

InputNode: will read data from a .data file and setup input nodes (in double types) according to data

HiddenLayer: This layer will be 2/3 the size of the length of each Input ArrayList

NeuralNetwork: This will be the main execution class, where I will setup a 2-d double array with randomized weights. Those weights will be multiplied by the input data to get a sum. There will be an error and a train method present here.

My question pertains to the InputNode and the HiddenLayer class. I want the InputNode class to read data from a .data file and setup each row from the .data into 13 input nodes. The last number of each row in the .data file represents a predictive number of 0-3: negative for something or above 3: positive for something

Example:

If I have 13 inputs, then I should have around 8 hidden nodes. Each hidden node will take in all the input nodes and multiply the input value of each input node times the weight of that specific hidden node and come up with either a 1 or -1 value.

How would I go about setting this up? Having some trouble with this since I had originally setup my NeuralNetwork class to setup an input ArrayList, but now have decided that it is better to create separate classes.

Here is what an example dataset would look like:

Data

Here is the code for my Node Class:

public class Node {

public Node(){
    
}

public void output(){
    //Don't know what to put here? Guess it can be modified for hiddenlayer and
        // Node classes 
}

}

Here is the code for my InputNode Class:

import java.io.File;
import java.io.FileNotFoundException;
import java.util.ArrayList;
import java.util.List;
import java.util.Scanner;

public class InputNode extends Node {

//declare arraylist for input nodes
private List inputs = new ArrayList();

//Constuctor for input nodes
public InputNode(File f){
    
    List tempInput = new ArrayList();
    this.inputs = tempInput;
    
    try {
        @SuppressWarnings("resource")
        Scanner inFile = new Scanner(f);
        
        //While there is another line in inFile.
        while (inFile.hasNextLine()){
            //Store that line into String line
            String line = inFile.nextLine();
            //Parition values separated by a comma
            String[] columns = line.split(",");
            /*System.out.println(columns.length); //Test code to see length of each column
            * code works and prints 14
            * */
            //create a double array of size columns
            double[] rows = new double[columns.length];
            for (int i = 0; i < columns.length; i++){
                //For each row...
                if (!columns[i].equals("?")){
                    //If the values in each row do not equal "?"
                    //Set rows[i] to the values in column[i]
                    rows[i] = Double.parseDouble(columns[i]);
                }
                else {
                    rows[i] = 0;
                }
            }
            inputs.add(rows);
            }
        
    } catch (FileNotFoundException e) {
        // TODO Auto-generated catch block
        e.printStackTrace();
    }
    
}




}

Trying to setup a Neural Network, having some trouble with structure

Answers (1)

Related Questions