PM> Install-Package Keras.NET -Version 3.8.4.4

除此之外，您将需要使用Windows CLI或Powershell中的pip安装程序来安装Keras和Tensorflow for Python：

pip install keraspip install tensorflow

它必须在Windows-CLI中看起来像：

C:\Users\YOURNAME>pip install keras

如果您对pip安装程序有疑问，请检查是否已将Python设置为系统环境。安装屏幕底部有一个复选框，称为“将Python添加到PATH”，这非常重要。

using System;namespace Keras.net_and_fashion_mnist
{class Program{static void Main(string[] args){KerasClass keras = new KerasClass();keras.TrainModel();}        }
}

KerasClass是更加有趣的：

using Keras.Datasets;
using Keras.Layers;
using Keras.Models;
using Keras.Utils;
using Numpy;
using System;
using System.IO;
using System.Linq;namespace Keras.net_and_fashion_mnist
{class KerasClass{public void TrainModel(){int batch_size = 1000;   //Size of the batches per epochint num_classes = 10;    //We got 10 outputs since //we can predict 10 different labels seen on the //dataset: https://github.com/zalandoresearch/fashion-mnist#labelsint epochs = 30;         //Amount on trainingperiods, //I figure it out that the maximum is something about //700 epochs, after this it won't increase the //accuracy siginificantly// input image dimensionsint img_rows = 28, img_cols = 28;// the data, split between train and test setsvar ((x_train, y_train), (x_test, y_test)) = FashionMNIST.LoadData(); // Load the datasets from // fashion MNIST, Keras.Net implement this directlyx_train.reshape(-1, img_rows, img_cols).astype(np.float32); //ByteArray needs //to be reshaped to fit the dimmensions of the y arraysy_train = Util.ToCategorical(y_train, num_classes); //here, you modify the //forecast data to 10 outputs//as we have 10 different //labels to predict (see the //Labels on the Dataset)y_test = Util.ToCategorical(y_test, num_classes);   //same for the test data //[hint: you can change this //in example you want to //make just a binary //crossentropy as you just //want to figure, i.e., if //this is a angleboot or notvar model = new Sequential();model.Add(new Dense(100, 784, "sigmoid")); //hidden dense layer, with 100 neurons, //you have 28*28 pixel which make //784 'inputs', and sigmoid function //as activationfunctionmodel.Add(new Dense(10, null, "sigmoid")); //Ouputlayer with 10 outputs,...model.Compile(optimizer: "sgd", loss: "categorical_crossentropy", metrics: new string[] { "accuracy" }); //we have a crossentropy as prediction //and want to see as well the //accuracy metric.var X_train = x_train.reshape(60000, 784); //this is actually very important. //C# works with pointers, //so if you have to reshape (again) //the function for the correct //processing, you need to write this //to a different varvar X_test = x_test.reshape(10000, 784);model.Fit(X_train, y_train, batch_size, epochs, 1); //now, we set the data to //the model with all the //arguments (x and y data, //batch size...the '1' is //just verbose=1Console.WriteLine("---------------------");Console.WriteLine(X_train.shape);Console.WriteLine(X_test.shape);Console.WriteLine(y_train[0]);Console.WriteLine(y_train[1]);       //some outputs...you can play with themvar y_train_pred = model.Predict(X_train);       //prediction on the train dataConsole.WriteLine(y_train_pred);model.Evaluate(X_test.reshape(-1, 784), y_test); //-1 tells the code that //it can figure out the size of //the array by itself}private byte[] openDatas(string path, int skip)      //just the open Data function. //As I mentioned, I did not work //with unzip stuff, you have //to unzip the data before //by yourself{var file = File.ReadAllBytes(path).Skip(skip).ToArray();return file;}//Hint: First, I was working by opening the data locally and //I wanted to figure it out how to present data to the arrays. //So you can use something like this and call this within the TrainModel() function://x_train = openDatas(@"PATH\OF\YOUR\DATAS\train-images-idx3-ubyte", 16);//y_train = openDatas(@"PATH\OF\YOUR\DATAS\train-labels-idx1-ubyte", 8);//x_test = openDatas(@"PATH\OF\YOUR\DATAS\t10k-images-idx3-ubyte", 16);//y_test = openDatas(@"PATH\OF\YOUR\DATAS\t10k-labels-idx1-ubyte", 8);}
}

兴趣点