pyspark 多层神经网络


from pyspark import SparkContext
from pyspark.sql import SQLContext
from pyspark.sql import SparkSession

from pyspark.ml.feature import StringIndexer, VectorAssembler
from pyspark.ml.classification import MultilayerPerceptronClassifier
from pyspark.ml.evaluation import MulticlassClassificationEvaluator
from pyspark.sql.types import DoubleType, IntegerType


if __name__ == "__main__":

    sc = SparkContext('local', 'mlp')
    sqlContext = SQLContext(sc)

    spark = SparkSession\
        .builder\
        .appName("MLPClassifier")\
        .getOrCreate()

    #read in csv as dataframe
    dataset = sqlContext.read.format('com.databricks.spark.csv').options(header='true').load('./data/titanic.csv')   
    dataset = dataset.drop('PassengerId','Name','Ticket','Cabin')       

    #set column types
    dataset = dataset.withColumn("Survived", dataset["Survived"].cast(IntegerType()))
    dataset = dataset.withColumn("Pclass", dataset["Pclass"].cast(IntegerType()))    
    dataset = dataset.withColumn("Age", dataset["Age"].cast(DoubleType()))
    dataset = dataset.withColumn("SibSp", dataset["SibSp"].cast(IntegerType()))
    dataset = dataset.withColumn("Parch", dataset["Parch"].cast(IntegerType()))
    dataset = dataset.withColumn("Fare", dataset["Fare"].cast(DoubleType()))

    #fill NaN
    avg_age = round(dataset.groupBy().avg("age").collect()[0][0],2)
    dataset = dataset.na.fill({'Age': avg_age})
    dataset = dataset.na.drop()

    #map categorical data
    indexer = StringIndexer(inputCol="Sex", outputCol="SexInd")
    dataset = indexer.fit(dataset).transform(dataset)

    indexer = StringIndexer(inputCol="Embarked", outputCol="EmbarkedInd")
    dataset = indexer.fit(dataset).transform(dataset)

    #assemble features
    assembler = VectorAssembler(
        inputCols=["Age","Pclass","SexInd","SibSp","Parch","Fare","EmbarkedInd"],
        outputCol="features")

    dataset = assembler.transform(dataset)

    (trainingData, testData) = dataset.randomSplit([0.8, 0.2])

    #MLP
    layers = [7, 8, 4, 2]  #input: 7 features; output: 2 classes
    mlp = MultilayerPerceptronClassifier(maxIter=100, layers=layers, labelCol="Survived", featuresCol="features", blockSize=128, seed=0)

    model = mlp.fit(trainingData)    
    result = model.transform(testData)

    prediction_label = result.select("prediction", "Survived")
    evaluator = MulticlassClassificationEvaluator(labelCol="Survived", predictionCol="prediction", metricName="accuracy")
    print ("MLP test accuracy: " + str(evaluator.evaluate(prediction_label)))

评论 2
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值