1. 定義資料結構
2. 讀取訓練資料
3. 選擇向量
4. 訓練模型
5. 預測

實現：

using System;
using System.Collections.Generic;
using System.Configuration;
using System.Text;
using Microsoft.ML;
using Microsoft.ML.Data;
using Microsoft.ML.Runtime.Api;
using Microsoft.ML.Trainers;
using Microsoft.ML.Transforms;

namespace MLNetLab
{
    // IrisData is used to provide training data, and as 
    // input for prediction operations
    // - First 4 properties are inputs/features used to predict the label
    // - Label is what you are predicting, and is only set when training
    public class IrisData
    {
        [Column("0")]
        public float SepalLength;

        [Column("1")]
        public float SepalWidth;

        [Column("2")]
        public float PetalLength;

        [Column("3")]
        public float PetalWidth;

        [Column("4")]
        [ColumnName("Label")]
        public string Label;
    }

    // IrisPrediction is the result returned from prediction operations
    public class IrisPrediction
    {
        [ColumnName("PredictedLabel")]
        public string PredictedLabels;
    }


    public class IrisRunner
    {
        private static string dataPath = ConfigurationManager.AppSettings["iris_file_name"];


        public static void Go()
        {
            // STEP 2: Create a pipeline and load your data
            var pipeline = new LearningPipeline();
            pipeline.Add(new TextLoader(dataPath).CreateFrom<IrisData>(separator: ','));

            // STEP 3: Transform your data
            // Assign numeric values to text in the "Label" column, because only
            // numbers can be processed during model training
            pipeline.Add(new Dictionarizer("Label"));

            // Puts all features into a vector
            pipeline.Add(new ColumnConcatenator("Features", "SepalLength", "SepalWidth", "PetalLength", "PetalWidth"));

            // STEP 4: Add learner
            // Add a learning algorithm to the pipeline. 
            // This is a classification scenario (What type of iris is this?)
            pipeline.Add(new StochasticDualCoordinateAscentClassifier());

            // Convert the Label back into original text (after converting to number in step 3)
            pipeline.Add(new PredictedLabelColumnOriginalValueConverter() { PredictedLabelColumn = "PredictedLabel" });

            // STEP 5: Train your model based on the data set
            var model = pipeline.Train<IrisData, IrisPrediction>();

            // STEP 6: Use your model to make a prediction
            // You can change these numbers to test different predictions
            var prediction = model.Predict(new IrisData()
            {
                SepalLength = 3.3f,
                SepalWidth = 1.6f,
                PetalLength = 0.2f,
                PetalWidth = 5.1f,
            });

            Console.WriteLine($"Predicted flower type is: {prediction.PredictedLabels}");
        }
    }
}
 

呼叫：

 static void Main(string[] args)
        {
            SpeechSynthesizer synthesizer = new SpeechSynthesizer();
            synthesizer.Volume = 100;  // 0...100
            synthesizer.Rate = -3;     // -10...10

            // Synchronous
            synthesizer.Speak("Hello , Microsoft");

            // Asynchronous
            //synthesizer.SpeakAsync("Hello World");


            Console.ReadLine();
        }