Accord.Net_API使用_SVM_的AND範例_使用_資料表示方法_筆記

首先到 Accord.Net 官網  做 API 下載

載點: http://accord-framework.net/index.html



這次要來學習 SVM 的範例

這是一個簡短範例改寫後的的執行畫面


顯示的是  做了 AND 的 結果

程式碼:

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;
using Accord.Controls;
using Accord.IO;
using Accord.MachineLearning.VectorMachines;
using Accord.MachineLearning.VectorMachines.Learning;
using Accord.Math;
using Accord.Statistics.Analysis;
using Accord.Statistics.Kernels;
using System.Data;

namespace SVM_example_ex
{
    class Program
    {
        static void Main(string[] args)
        {
            svm_input();
        }

        private static void svm_input()
        {            
            double[][] problem =
            {
                //             a    b    a & b
                new double[] { 0,   0,     0    },
                new double[] { 0,   1,     0    },
                new double[] { 1,   0,     0    },
                new double[] { 1,   1,     1    },
            };
            //從矩陣中取得 column vector
            //double[][] inputs;
            double[][] inputs = problem.GetColumns(0, 1);
            
            //取得double[][] problem 浮點數二維陣列取到第二直行的數值,為一個vector = [0,0,0,1]
            int[] outputs = problem.GetColumn(2).ToInt32();
            for(int i = 0; i<=3;i++)
                Console.WriteLine("第{0}個存在outputs的元素:{1}",i,outputs[i]);

            for(int j=0;j<=3;j++)
                for(int k=0;k<=2;k++)
                {
                    Console.WriteLine("第({0},{1})個元素為:{2}",j,k,problem[j][k]);
                }
            ScatterplotBox.Show("AND", inputs, outputs).Hold();
            // However, SVMs expect the output value to be
            // either -1 or +1. As such, we have to convert
            // it so the vector contains { -1, -1, -1, +1 }://資料的分類標記
            outputs = outputs.Apply(x => x == 0 ? -1 : 1);
            // Create a new linear-SVM for two inputs (a and b)
            SupportVectorMachine svm = new SupportVectorMachine(2);

            // Create a L2-regularized L2-loss support vector classification
            var teacher = new LinearDualCoordinateDescent(svm, inputs, outputs)
            {
                Loss = Loss.L2, //
                Complexity = 1000, //複雜度
                Tolerance = 1e-5 //表示10的-5次方
            };

            // Learn the machine
            double error = teacher.Run(computeError: true);

            // Compute the machine's answers for the learned inputs
            int[] answers = inputs.Apply(x => Math.Sign(svm.Compute(x)));
            //Math.Sign

            // Plot the results
            ScatterplotBox.Show("SVM's answer", inputs, answers).Hold();
        }
    }
}