import java.util.Random; 
 
class Example{
    double input[];
    double output;
 
    Example(double[] in, double out){
        this.input = in;
        this.output = out;
    } 
}
 
class HiddenSigmoidUnit{
    double output, error;
    double input[];
    double w[];  
    double n = 0.05;  
    double bias = 1;
 
    static Random rand = new Random();
    HiddenSigmoidUnit(){ 
        w = new double[2];   
        int s = rand.nextBoolean()?1:-1;
        for(int i=0;i<w.length;++i)w[i] = s*rand.nextDouble();  
    } 
 
    public double computeOutput(double[] input){
        this.input = input;
        output = bias+w[0]*input[0] + w[1]*input[1];
        output = 1/(1+Math.pow(Math.E, -output));    
        return output;
    }
 
    public double computeError(double w, double outputUnitError){
        error = (output)*(1-output)*(outputUnitError*w);  
        return error;        
    }
 
    public void fixError(){  
        for(int i=0;i<input.length;++i) w[i] += n*error*input[i]; 
    }
}
 
class OutputUnit{
    double t, output, error;
    double input[];
    double w[];  
    double bias=1;
    double n = 0.05; 
 
    static Random rand = new Random();
 
    OutputUnit(int hUnits){ 
        w = new double[hUnits];
        int s = rand.nextBoolean()?1:-1;
        //Weights between -1 to 1
        for(int i=0;i<w.length;++i)w[i] = s*rand.nextDouble();
    } 
 
    public void computeOutput(double[] input) { 
        this.input = input;  
        output = bias+input[0]*w[0]+input[1]*w[1];
        output = 1/(1+Math.pow(Math.E, -output));   
    } 
 
    public void computeError(double t){ 
        this.t = t; 
        error = (output)*(1-output)*(t-output); 
    }
 
    public void fixError() {     
        for(int i=0;i<w.length;++i) w[i] += n*error*input[i]; 
    }   
} 
 
class Net{
    static Random rand = new Random();
    static OutputUnit outputUnit;
    static HiddenSigmoidUnit[] hlayer;
 
    static Example[] examples = {
            new Example(new double[]{0, 0}, 0),
            new Example(new double[]{0,  1}, 1),
            new Example(new double[]{1, 0}, 1),
            new Example(new double[]{1,  1}, 0)
    }; 
 
    public static void main(String[] args){  
        //units in hidden layer
        int nhidden = 2;
 
        //Holds hidden layers' units
        hlayer = new HiddenSigmoidUnit[nhidden];
 
        //Initialize
        for(int i=0;i<hlayer.length;++i)hlayer[i] = new HiddenSigmoidUnit();
        outputUnit = new OutputUnit(hlayer.length);
 
        //test multiple times for convergence to optimal values
        for(int check=0;check<100;++check){
 
            for(int i=0;i<hlayer.length;++i)hlayer[i] = new HiddenSigmoidUnit();
            outputUnit = new OutputUnit(hlayer.length);
 
            //iteration count
            for(int iteration = 0;;++iteration){ 
 
                //if count == 4, all the examples were classified correctly, break
                int count=0;
 
                //Shuffle examples
                /*
                for(int i=0;i<examples.length;++i){
                    for(int j=i+1;j<examples.length;++j){
                        int choose = rand.nextInt(j);
                        Example t = examples[choose];
                        examples[choose] = examples[j];
                        examples[j] = t;
                    }
                }
                */
 
                //for each example
                for(int i=0;i<examples.length;++i){
 
                    Example example = examples[i]; 
 
                    //collect the outputs from hidden layer to pass on to output unit
                    double[] outputsHLayer = new double[hlayer.length]; 
                    for(int j=0;j<hlayer.length;++j) outputsHLayer[j] = hlayer[j].computeOutput(example.input); 
 
                    //pass to output unit
                    outputUnit.computeOutput(outputsHLayer);
 
                    //Compute Errors for output and hidden layer units
                    //Passing the true output, compute outputUnit error
                    outputUnit.computeError(example.output);   
 
                    //compute hidden layer - units' error
                    for(int j=0;j<hlayer.length;++j) hlayer[j].computeError(outputUnit.w[j], outputUnit.error); 
 
                    //fix errors
                    outputUnit.fixError();
                    for(int j=0;j<hlayer.length;++j) hlayer[j].fixError();  
 
                    //check if output units result matches the current example
                    int o = outputUnit.output>0.5?1:0;
                    if(o==example.output)count++;
                } //end-for each example
 
                if(iteration>90000){
                    //Probably not going to converge to optimal values
                    System.out.println("\nIterations > 90000, stop...");
                    displayOutputs(); 
                    break;
                }
 
                //if it matches all examples, stop training
                if(count==examples.length){
                    System.out.println("\nTraining complete. No of iterations = "+iteration);
                    displayOutputs();
                    break;
                }
            } //end for-each 
        }//end for-each re-initialization
    }//end main
 
    //Apply the learned weights to all the examples
    public static void displayOutputs(){
        System.out.println("Displaying outputs for all examples... ");
        double[] outputsHLayer = new double[hlayer.length]; 
 
        for(int e=0;e<examples.length;++e){
            for(int j=0;j<hlayer.length;++j) {
                outputsHLayer[j] = hlayer[j].computeOutput(examples[e].input);
            } 
            outputUnit.computeOutput(outputsHLayer); 
            System.out.println(outputUnit.output);
        }
    }
}

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