Deep learning sample

[Supriya Das]

Hello all,

    I am using seep learning sample for digit recognition. In the sample application when ever i run the application each and every time it is training the data. I want to is it possible to train data only once and the i can test multiple number of times. Please give a suggestion to how to solve my problem? code is as follow,

using Accord.Neuro;

using Accord.Neuro.ActivationFunctions;

using Accord.Neuro.Learning;

using Accord.Neuro.Networks;

using Accord.Math;

using System;

using System.Collections.Generic;

using System.Linq;

using System.Text;

using System.Threading.Tasks;

using AForge.Neuro.Learning;

using System.IO;

namespace DeepLearning

{

    class Program

    {

        static void Main(string[] args)

        {

            double[][] inputs;

            double[][] outputs;

            double[][] testInputs;

            double[][] testOutputs;

            // Load ascii digits dataset.

            inputs = DataManager.LoadMultiClass(@"../../../data/Train.txt", out outputs);

            testInputs = DataManager.Load(@"../../../data/Test.txt", out testOutputs);

            // Setup the deep belief network and initialize with random weights.

            int inputLength = inputs.First().Length;

            DeepBeliefNetwork network = new DeepBeliefNetwork(inputs.First().Length, 68, 68);

            new GaussianWeights(network, 0.1).Randomize();

            network.UpdateVisibleWeights();

            

            // Setup the learning algorithm.

            DeepBeliefNetworkLearning teacher = new DeepBeliefNetworkLearning(network)

            {

                Algorithm = (h, v, i) => new ContrastiveDivergenceLearning(h, v)

                {

                    LearningRate = 0.1,

                    Momentum = 0.5,

                    Decay = 0.001,

                }

            };

            // Setup batches of input for learning.

            int batchCount = Math.Max(1, inputs.Length / 100);

            // Create mini-batches to speed learning.

            int[] groups = Accord.Statistics.Tools.RandomGroups(inputs.Length, batchCount);

            double[][][] batches = inputs.Subgroups(groups);

            // Learning data for the specified layer.

            double[][][] layerData;

            // Unsupervised learning on each hidden layer, except for the output layer.

            for (int layerIndex = 0; layerIndex < network.Machines.Count - 1; layerIndex++)

            {

                teacher.LayerIndex = layerIndex;

                layerData = teacher.GetLayerInput(batches);

                for (int i = 0; i < 200; i++)

                {

                    double error = teacher.RunEpoch(layerData) / inputs.Length;

                    if (i % 10 == 0)

                    {

                        Console.WriteLine(i + ", Error = " + error);

                    }

                }

            }

            // Supervised learning on entire network, to provide output classification.

            var teacher2 = new BackPropagationLearning(network)

            {

                LearningRate = 0.1,

                Momentum = 0.5

            };

            // Run supervised learning.

            for (int i = 0; i < 500; i++)

            {

                double error = teacher2.RunEpoch(inputs, outputs) / inputs.Length;

                if (i % 10 == 0)

                {

                    Console.WriteLine(i + ", Error = " + error);

                }

            }

            // Test the resulting accuracy.

            int correct = 0;

            for (int i = 0; i < testInputs.Length; i++)

            {

                double[] outputValues = network.Compute(testInputs[i]);

                if (DataManager.FormatOutputResult(outputValues) == DataManager.FormatOutputResult(testOutputs[i]))

                {

                    correct++;

                }

            }

            Console.WriteLine("Correct " + correct + "/" + testInputs.Length + ", " + Math.Round(((double)correct / (double)testInputs.Length * 100), 2) + "%");

            Console.Write("Press any key to quit ..");

            Console.ReadKey();

        }

    }

}