基于Java实现CNN，并附MNIST和语音（MFCC特征）性别识别示例。

构建CNN LayerBuilder builder = new LayerBuilder(); builder.addLayer(Layer.buildInputLayer(new Size(28, 28))); builder.addLayer(Layer.buildConvLayer(6, new Size(5, 5))); builder.addLayer(Layer.buildSampLayer(new Size(2, 2))); builder.addLayer(Layer.buildConvLayer(12, new Size(5, 5))); builder.addLayer(Layer.buildSampLayer(new Size(2, 2))); builder.addLayer(Layer.buildOutputLayer(10)); CNN cnn = new CNN(builder, 50); 运行MNIST数据集 String fileName = "data/train.format"; Dataset dataset = Dataset.load(fileName, ",", 784); cnn.train(dat aset, 100); Dataset testset = Dataset.load("data/test.format", ",", -1); cnn.predict(testset, "data/test.predict"); 计算精度可以达到97.8%。