线性回归
1.简介
线性回归是利用称为线性回归方程的函数对一个或多个自变量和
因变量之间关系进行建模的各种回归分析
只有一个自变量的情况称为简单回归,大于一个自变量情况的叫做
多元回归
2.基本过程
构造预测函数(h函数)
•构造cost函数
•利用梯度下降(上升)法来计算 的最小值
•梯度下降(上升)计算的向量化
3.基本理论
h函数:
cost函数:
梯度下降:
Mean Square Error:
4.程序示例
import org.apache.spark.SparkContext
import org.apache.spark.mllib.regression.LinearRegressionWithSGD
import org.apache.spark.mllib.regression.LabeledPoint
/**
* Created by shaokai on 14-9-12.
*/
object LinearRegression {
def main(args : Array[String]){
val sc = new SparkContext("local[2]","BinaryClassification","/Users/software/spark-0.9.0-incubating-bin-hadoop1")
val data = sc.textFile("/Users/workspace/chinahadoop/data/ridge-data/lpsa.data")
val parsedData = data.map { line =>
val parts = line.split(',')
LabeledPoint(parts(0).toDouble, parts(1).split(' ').map(x => x.toDouble).toArray)
}
//构建模型
val numIterations = 20
val model = LinearRegressionWithSGD.train(parsedData, numIterations)
//预测
val valuesAndPreds = parsedData.map { point =>
val prediction = model.predict(point.features)
(point.label, prediction)
}
//计算MSE
val MSE = valuesAndPreds.map{ case(v, p) => math.pow((v - p), 2)}.reduce(_ + _)/valuesAndPreds.count
println("training Mean Squared Error = " + MSE)
}
}