C# 机器学习实战SVM多分类

1：先用一下  Accord.MachineLearning(在NuGet下载即可)

2：引用

using Accord.MachineLearning.VectorMachines.Learning;
using Accord.Statistics.Kernels;

3：实例SVM多分类

Code:

double[][] inputs =
            {
                new double[] { 0 },
                new double[] { 3 },
                new double[] { 1 },
                new double[] { 2 },
            };

            double[][] inputsTest =
            {
                new double[] { 4 },
            };

            // Outputs for each of the inputs
            int[] outputs =
            {
                0,
                3,
                1,
                2,
            };

            // Create the Multi-label learning algorithm for the machine
            var teacher = new MulticlassSupportVectorLearning<Linear>()
            {
                Learner = (p) => new SequentialMinimalOptimization<Linear>()
                {
                    Complexity = 10000.0 // Create a hard SVM
                }
            };

            // Learn a multi-label SVM using the teacher
            var svm = teacher.Learn(inputs, outputs);

            // Compute the machine answers for the inputs
            int[] answers = svm.Decide(inputsTest);

            Console.WriteLine(answers[0]);
            Console.ReadKey();

github网址:    https://github.com/accord-net/framework/wiki/Classification