Sunday, January 5, 2014

Simple C# Artificial Neural Network

Artificial intelligence interests me, especially when modeled from or inspired by biology. Artificial neural networks are one such example:

"Artificial neural networks are computational models inspired by animals' central nervous systems (in particular the brain) that are capable of machine learning and pattern recognition. They are usually presented as systems of interconnected "neurons" that can compute values from inputs by feeding information through the network." - Wikipedia

I'm not going to try explain how artificial neural networks work. The topic is much too complex and there is already good documentation available elsewhere. If you are interested in learning more about artificial neural networks then I recommend the following article: Neural Network Back-Propagation Using C#. Most of the code in this post is based on the ideas presented in that article.

There are many different types of artificial neural networks. My implementation is a feed-forward, multi-layer, perceptron network which uses the back-propagation algorithm for learning. I believe this is one of the most common types of neural networks and is a good place to start for beginners. I'm still a beginner myself; I understand the basic concepts but the math is still slightly mysterious. Hopefully I can find time in the future to learn more about neural networks and artificial intelligence.

I tried to make the code as simple as possible. Code is below:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;

namespace NeuralNetwork
{
    class Program
    {
        static void Main(string[] args)
        {
            var network = new NeuralNetwork(2, 3, 1);

            Console.WriteLine("Training Network...");
            for (int i = 0; i < 100000; i++)
            {
                network.Train(0, 0);
                network.BackPropagate(1);

                network.Train(1, 0);
                network.BackPropagate(0);

                network.Train(0, 1);
                network.BackPropagate(0);

                network.Train(1, 1);
                network.BackPropagate(1);
            }

            double error;
            double output;

            output = network.Compute(0, 0)[0];
            error = network.CalculateError(1);
            Console.WriteLine("0 XOR 0 = " + output.ToString("F5") + ", Error = " + error.ToString("F5"));

            output = network.Compute(1, 0)[0];
            error = network.CalculateError(0);
            Console.WriteLine("1 XOR 0 = " + output.ToString("F5") + ", Error = " + error.ToString("F5"));

            output = network.Compute(0, 1)[0];
            error = network.CalculateError(0);
            Console.WriteLine("0 XOR 1 = " + output.ToString("F5") + ", Error = " + error.ToString("F5"));

            output = network.Compute(1, 1)[0];
            error = network.CalculateError(1);
            Console.WriteLine("1 XOR 1 = " + output.ToString("F5") + ", Error = " + error.ToString("F5"));
        }
    }

    public class NeuralNetwork
    {
        public double LearnRate { get; set; }
        public double Momentum { get; set; }
        public List<Neuron> InputLayer { get; set; }
        public List<Neuron> HiddenLayer { get; set; }
        public List<Neuron> OutputLayer { get; set; }
        static Random random = new Random();

        public NeuralNetwork(int inputSize, int hiddenSize, int outputSize)
        {
            LearnRate = .9;
            Momentum = .04;
            InputLayer = new List<Neuron>();
            HiddenLayer = new List<Neuron>();
            OutputLayer = new List<Neuron>();

            for (int i = 0; i < inputSize; i++)
                InputLayer.Add(new Neuron());

            for (int i = 0; i < hiddenSize; i++)
                HiddenLayer.Add(new Neuron(InputLayer));

            for (int i = 0; i < outputSize; i++)
                OutputLayer.Add(new Neuron(HiddenLayer));
        }

        public void Train(params double[] inputs)
        {
            int i = 0;
            InputLayer.ForEach(a => a.Value = inputs[i++]);
            HiddenLayer.ForEach(a => a.CalculateValue());
            OutputLayer.ForEach(a => a.CalculateValue());
        }

        public double[] Compute(params double[] inputs)
        {
            Train(inputs);
            return OutputLayer.Select(a => a.Value).ToArray();
        }

        public double CalculateError(params double[] targets)
        {
            int i = 0;
            return OutputLayer.Sum(a => Math.Abs(a.CalculateError(targets[i++])));
        }

        public void BackPropagate(params double[] targets)
        {
            int i = 0;
            OutputLayer.ForEach(a => a.CalculateGradient(targets[i++]));
            HiddenLayer.ForEach(a => a.CalculateGradient());
            HiddenLayer.ForEach(a => a.UpdateWeights(LearnRate, Momentum));
            OutputLayer.ForEach(a => a.UpdateWeights(LearnRate, Momentum));
        }

        public static double NextRandom()
        {
            return 2 * random.NextDouble() - 1;
        }

        public static double SigmoidFunction(double x)
        {
            if (x < -45.0) return 0.0;
            else if (x > 45.0) return 1.0;
            return 1.0 / (1.0 + Math.Exp(-x));
        }

        public static double SigmoidDerivative(double f)
        {
            return f * (1 - f);
        }
    }

    public class Neuron
    {
        public List<Synapse> InputSynapses { get; set; }
        public List<Synapse> OutputSynapses { get; set; }
        public double Bias { get; set; }
        public double BiasDelta { get; set; }
        public double Gradient { get; set; }
        public double Value { get; set; }

        public Neuron()
        {
            InputSynapses = new List<Synapse>();
            OutputSynapses = new List<Synapse>();
            Bias = NeuralNetwork.NextRandom();
        }

        public Neuron(List<Neuron> inputNeurons) : this()
        {
            foreach (var inputNeuron in inputNeurons)
            {
                var synapse = new Synapse(inputNeuron, this);
                inputNeuron.OutputSynapses.Add(synapse);
                InputSynapses.Add(synapse);
            }
        }

        public virtual double CalculateValue()
        {
            return Value = NeuralNetwork.SigmoidFunction(InputSynapses.Sum(a => a.Weight * a.InputNeuron.Value) + Bias);
        }

        public virtual double CalculateDerivative()
        {
            return NeuralNetwork.SigmoidDerivative(Value);
        }

        public double CalculateError(double target)
        {
            return target - Value;
        }

        public double CalculateGradient(double target)
        {
            return Gradient = CalculateError(target) * CalculateDerivative();
        }

        public double CalculateGradient()
        {
            return Gradient = OutputSynapses.Sum(a => a.OutputNeuron.Gradient * a.Weight) * CalculateDerivative();
        }

        public void UpdateWeights(double learnRate, double momentum)
        {
            var prevDelta = BiasDelta;
            BiasDelta = learnRate * Gradient; // * 1
            Bias += BiasDelta + momentum * prevDelta;

            foreach (var s in InputSynapses)
            {
                prevDelta = s.WeightDelta;
                s.WeightDelta = learnRate * Gradient * s.InputNeuron.Value;
                s.Weight += s.WeightDelta + momentum * prevDelta;
            }
        }
    }

    public class Synapse
    {
        public Neuron InputNeuron { get; set; }
        public Neuron OutputNeuron { get; set; }
        public double Weight { get; set; }
        public double WeightDelta { get; set; }
        
        public Synapse(Neuron inputNeuron, Neuron outputNeuron)
        {
            InputNeuron = inputNeuron;
            OutputNeuron = outputNeuron;
            Weight = NeuralNetwork.NextRandom();
        }
    }
}