签到
- 苹果/安卓/wp
- 苹果/安卓/wp
客户端
0.0

0.00

经管百科

人大经济论坛 › 论坛 › 数据科学与人工智能 › IT基础 › Scala及其他JVM语言 › 【GitHub】Spark Scala Learning Note

楼主: ReneeBK

3318 26

【GitHub】Spark Scala Learning Note [推广有奖]

21楼

ReneeBK 发表于 2017-2-21 09:29:57

Scaling
Making features have approximately zero mean by replacing each field x with x-m, and values within an unit standard deviation by dividing the range of feature.+
Import
import org.apache.spark.mllib.feature.StandardScaler
Code
val scaler = new StandardScaler(withMean = true, withStd = true).fit(trainingSet.map(dp => dp.features))
Scale the training and test set.
val scaledTrainingSet = trainingSet.map(dp => new LabeledPoint(dp.label, scaler.transform(dp.features))).cache()
val scaledTestSet = testSet.map(dp => new LabeledPoint(dp.label, scaler.transform(dp.features))).cache()

复制代码

回复

22楼

ReneeBK 发表于 2017-2-21 09:30:58

Training the Model
Import
import org.apache.spark.mllib.regression.LinearRegressionWithSGD
Code
val regression=new LinearRegressionWithSGD().setIntercept(true)
regression.optimizer.setNumIterations(1000).setStepSize(0.1)
//Let's create a model out of our training examples.
val model=regression.run(scaledTrainingSet)

复制代码

回复

23楼

franky_sas 发表于 2017-2-21 11:16:08

Thanks for sharing.

回复

24楼

钱学森64 发表于 2017-2-21 12:38:04

谢谢分享

回复

25楼

kkkm_db 发表于 2017-2-21 16:14:27

谢谢分享！

回复

26楼

在职认证

发表于 2017-2-21 18:41:54 来自手机

ReneeBK 发表于 2017-2-21 07:35
**** 本内容被作者隐藏 ****

学习学习

回复

27楼

钱学森64 发表于 2017-2-21 19:13:06

谢谢分享

回复

发帖

本版微信群

jg-xs1
拉您进交流群

京ICP备16021002号-2 京B2-20170662号京公网安备 11010802022788号论坛法律顾问：王进律师知识产权保护声明免责及隐私声明