多层感知机（MLP）算法原理及Spark MLlib调用实例（Scala/Java/Python）

多层感知机

算法简介：

        多层感知机是基于反向人工神经网络（feedforwardartificial neural network）。多层感知机含有多层节点，每层节点与网络的下一层节点完全连接。输入层的节点代表输入数据，其他层的节点通过将输入数据与层上节点的权重w以及偏差b线性组合且应用一个激活函数，得到该层输出。多层感知机通过方向传播来学习模型，其中我们使用逻辑损失函数以及L-BFGS。K＋1层多层感知机分类器可以写成矩阵形式如下：

 

中间层节点使用sigmoid方程：

 

输出层使用softmax方程：

 

输出层中N代表类别数目。

参数：

featuresCol:

类型：字符串型。

含义：特征列名。

labelCol:

类型：字符串型。

含义：标签列名。

layers:

类型：整数数组型。

含义：层规模，包括输入规模以及输出规模。

maxIter:

类型：整数型。

含义：迭代次数（>=0）。

predictionCol:

类型：字符串型。

含义：预测结果列名。

seed:

类型：长整型。

含义：随机种子。

stepSize:

类型：双精度型。

含义：每次迭代优化步长。

tol:

类型：双精度型。

含义：迭代算法的收敛性。

示例：

Scala:

import org.apache.spark.ml.classification.MultilayerPerceptronClassifierimport org.apache.spark.ml.evaluation.MulticlassClassificationEvaluator// Load the data stored in LIBSVM format as a DataFrame.val data = spark.read.format("libsvm")  .load("data/mllib/sample_multiclass_classification_data.txt")// Split the data into train and testval splits = data.randomSplit(Array(0.6, 0.4), seed = 1234L)val train = splits(0)val test = splits(1)// specify layers for the neural network:// input layer of size 4 (features), two intermediate of size 5 and 4// and output of size 3 (classes)val layers = Array[Int](4, 5, 4, 3)// create the trainer and set its parametersval trainer = new MultilayerPerceptronClassifier()  .setLayers(layers)  .setBlockSize(128)  .setSeed(1234L)  .setMaxIter(100)// train the modelval model = trainer.fit(train)// compute accuracy on the test setval result = model.transform(test)val predictionAndLabels = result.select("prediction", "label")val evaluator = new MulticlassClassificationEvaluator()  .setMetricName("accuracy")println("Accuracy: " + evaluator.evaluate(predictionAndLabels))

Java:

import org.apache.spark.sql.Dataset;import org.apache.spark.sql.Row;import org.apache.spark.sql.SparkSession;import org.apache.spark.ml.classification.MultilayerPerceptronClassificationModel;import org.apache.spark.ml.classification.MultilayerPerceptronClassifier;import org.apache.spark.ml.evaluation.MulticlassClassificationEvaluator;// Load training dataString path = "data/mllib/sample_multiclass_classification_data.txt";Dataset<Row> dataFrame = spark.read().format("libsvm").load(path);// Split the data into train and testDataset<Row>[] splits = dataFrame.randomSplit(new double[]{0.6, 0.4}, 1234L);Dataset<Row> train = splits[0];Dataset<Row> test = splits[1];// specify layers for the neural network:// input layer of size 4 (features), two intermediate of size 5 and 4// and output of size 3 (classes)int[] layers = new int[] {4, 5, 4, 3};// create the trainer and set its parametersMultilayerPerceptronClassifier trainer = new MultilayerPerceptronClassifier()  .setLayers(layers)  .setBlockSize(128)  .setSeed(1234L)  .setMaxIter(100);// train the modelMultilayerPerceptronClassificationModel model = trainer.fit(train);// compute accuracy on the test setDataset<Row> result = model.transform(test);Dataset<Row> predictionAndLabels = result.select("prediction", "label");MulticlassClassificationEvaluator evaluator = new MulticlassClassificationEvaluator()  .setMetricName("accuracy");System.out.println("Accuracy = " + evaluator.evaluate(predictionAndLabels));

Python：

from pyspark.ml.classification import MultilayerPerceptronClassifierfrom pyspark.ml.evaluation import MulticlassClassificationEvaluator# Load training datadata = spark.read.format("libsvm")\    .load("data/mllib/sample_multiclass_classification_data.txt")# Split the data into train and testsplits = data.randomSplit([0.6, 0.4], 1234)train = splits[0]test = splits[1]# specify layers for the neural network:# input layer of size 4 (features), two intermediate of size 5 and 4# and output of size 3 (classes)layers = [4, 5, 4, 3]# create the trainer and set its parameterstrainer = MultilayerPerceptronClassifier(maxIter=100, layers=layers, blockSize=128, seed=1234)# train the modelmodel = trainer.fit(train)# compute accuracy on the test setresult = model.transform(test)predictionAndLabels = result.select("prediction", "label")evaluator = MulticlassClassificationEvaluator(metricName="accuracy")print("Accuracy: " + str(evaluator.evaluate(predictionAndLabels)))