标签：eid 去重 rank 显示 seq off contain 训练 改进
搜索推荐，主要有以下几种形式：一、根据人口统计学推荐：此推荐方式需要建立用户模型，并且需要获取用户的具体信息，然后根据矩阵运算，计算相似度，此方式最大缺陷是获取用户的隐私，应用不多；
二、基于内容的推荐：根据特定用户的历史数据，推荐出相似的产品。缺点是需要建立item model,比较费时。
三、基于协同过滤，是目前搜索推荐中应用最广泛的，不需要建立item model,省事，效果比较好。协同过滤的本质，可以概括为"物以类聚，人以群分"，分别指基于物品的协同过滤和基于user的协同过滤。还有基于机器学习的协同过滤，总共这三种形式。第一种的优点是没有冷启动问题，基于用户历史行为的推荐，有冷启动问题。
亚马逊是搜索推荐的鼻祖，把搜索推荐运用到了极致，主要有以下形式：一，基于内容的推荐，主要有：①每日新产品的推荐；②基于用户历史数据的推荐（比如猜你喜欢）；二，基于协同过滤的推荐：①根据KNN算法，计算相似用户，推荐其他用户群喜欢的产品（去重）②根据FP-Growth model进行相关度挖掘，捆绑销售③基于ALS算法，推荐产品。关于ALS算法，有一篇经典的博文，是spark MLlib的源码贡献者之一写的，很深入。地址：http://www.csdn.net/article/2015-05-07/2824641 认真研究几遍，会有很多收获。本文主要讲ALS应用，所以会比较简单，但是，想成为这个领域的佼佼者，一定要深入研究理论，寻求突破和创新，切不可一辈子做一个低级的码奴。废话不说了，上代码：
package com.txq.spark.test
import java.io.File
import org.apache.log4j.{Level, Logger}
import org.apache.spark.mllib.recommendation.{ALS, MatrixFactorizationModel, Rating}
import org.apache.spark.rdd.RDD
import org.apache.spark.{SparkConf, SparkContext}
import scala.io.Source
/**
* ALS矩阵分解搜索推荐：
* 两个数据源：①rating.dat ----->userid itemid rate
* ②items.dat----itemid->item(Map)
* 思路：1.实现数据的partitions,数据格式为(key,value)形式,key为时间戳，对10求余数，根据余数分区，提高计算的
* 并行度;
* 2.取上述数据的values部分，训练出一个最优的model出来，评判标准为MSE;
* 3.根据最优model对特定用户推荐产品，注意去除该用户已经评分的产品.
*/
object MovieLensALS {
System.setProperty("hadoop.home.dir", "D://hadoop-2.6.2")
val userId = 0//被推荐用户的id号
def main(args: Array[String]): Unit = {
//屏蔽不必要的日志显示在终端上
Logger.getLogger("org.apache.spark").setLevel(Level.WARN)
Logger.getLogger("org.eclipse.jetty.server").setLevel(Level.OFF)
if(args.length != 2){
println("Usage:/path/to/spark/bin/spark-submit --driver-memory 2g --class " +
"week6.MovieLensALS" + "target/scala-*/movielens-als-ssembly-*.jar movieLensHomeDir personalRatingsFile")
sys.exit(1)
}
//设置运行环境
val conf = new SparkConf().setAppName("MovieLensALS")
val sc = new SparkContext(conf)
//装载用户评分，该评分由评分器生成
val myRatings = loadRatings(args(1))
val myRatingsRDD = sc.parallelize(myRatings,1)
//样本数据目录
val movieLensHomeDir = args(0)
//装载样本评分数据，其中最后一列Timestamp取除10的余数作为key，Rating为值，即(Int,Rating)
val ratings = sc.textFile(new File(movieLensHomeDir,"ratings.dat").toString).map{ line =>
val fields = line.split("::")
(fields(3).toLong % 10,Rating(fields(0).toInt,fields(1).toInt,fields(2).toDouble))
}
//装载电影目录对照表（电影ID->电影标题）
val movies = sc.textFile(new File(movieLensHomeDir,"movies.dat").toString).map{line =>
val fields = line.split("::")
(fields(0).toInt,fields(1))
}.collect().toMap
val numRatings = ratings.count()
val numUsers = ratings.map(_._2.user).distinct().count()
val numMovies = ratings.map(_._2.product).distinct().count()
println("Got " + numRatings + " ratings from " + numUsers + " users on" + numMovies + " movies.")
//将样本评分表以key值切分成3个部分，分别用于训练（60%，并加入用户评分），校验（20%），测试（20%）
//该数据在计算过程中要多次应用到，所以cache到内存
val numPartitions = 4
val training = ratings.filter(x => x._1 < 6).values.union(myRatingsRDD).repartition(numPartitions).cache()
val test = ratings.filter(x => x._1 >= 8).values.repartition(numPartitions).cache()
val validation = ratings.filter(x => x._1>= 6 && x._1 < 8).values.repartition(numPartitions).cache()
val numTraining = training.count()
val numValidation = validation.count()
val numTest = test.count()
println("Training: "+ numTraining + ",validation: " + numValidation + ",test: " + numTest)
//训练不同参数下的模型,并在校验集中校验，获取最佳参数下的模型
val ranks = List(8,9)
val lambdas = List(0.1,10.0)
val numIters = List(10,20)
var bestModel:Option[MatrixFactorizationModel] = None
var bestValidationRmse = Double.MaxValue
var bestRank = 0
var bestLambda = -1.0
var bestNumIter = -1
for(rank <- ranks;lambda <- lambdas;numIter <- numIters){
val model = ALS.train(training,rank,numIter,lambda)
val validationRmse = computeRmse(model,validation)
println("RMSE(validation) = " + validationRmse + " for the model trained with rank = " + rank + " with lambda = " + lambda + " with numIterations = " + numIter)
if(validationRmse < bestValidationRmse){
bestModel = Some(model)
bestValidationRmse = validationRmse
bestRank = rank
bestLambda = lambda
bestNumIter = numIter
}
}
//用最佳模型预测测试集的评分，并计算和实际评分之间的均方根误差
val testRmse = computeRmse(bestModel.get,test)
println("The best model was trained with rank = " + bestRank + " with lambda = " + bestLambda + " with best NumIterations = " + bestNumIter)
//create a naive baseline and compare it with the best model
val meanRating = training.union(validation).map(_.rating).mean()
val baselineRmse = math.sqrt(test.map(x => math.pow(meanRating - x.rating,2)).mean())
val improvement =(testRmse-baselineRmse) / baselineRmse * 100
println("The best model improves the base line by " + "%1.2f".format(improvement + "%."))
//推荐前十部最感兴趣的电影，注意要剔除用户已经评分的电影
val myRatedMovieIds = myRatings.map(_.product).toSet
val candidates = sc.parallelize(movies.keys.filter(!myRatedMovieIds.contains(_)).toSeq)
val recommendations = bestModel.get.predict(candidates.map((userId,_))).collect().sortBy(-_.rating).take(10)
var i = 1
println("Movies recommended for you: ")
recommendations.foreach{ r =>
println("%2d".format(i) + ": " + movies(r.product))
i += 1
}
sc.stop()
}
/**装载用户评分文件 **/
def loadRatings(path:String):Seq[Rating] = {
val lines = Source.fromFile(path).getLines()
val ratings = lines.map{line =>
val fields = line.split("::")
Rating(fields(0).toInt,fields(1).toInt,fields(2).toDouble)
}.filter(_.rating > 0.0)
if(ratings.isEmpty){
sys.error("No ratings provided.")
} else {
ratings.toSeq
}
}
/** 计算**/
def computeRmse(model:MatrixFactorizationModel,data:RDD[Rating]):Double = {
val precisions:RDD[Rating] = model.predict(data.map(x => (x.user,x.product)))
val precisionsAndReal = precisions.map(x => ((x.user,x.product),x.rating)).join(data.map(line =>((line.user,line.product
),line.rating))).values
math.sqrt(precisionsAndReal.map(x => math.pow(x._1 - x._2,2)).mean())
}
}
佟氏出品，必属精品。坚持算法和数据挖掘领域的研究，不断改进代码的质量，跻身一流程序员行列！
标签：eid 去重 rank 显示 seq off contain 训练 改进
原文地址：http://www.cnblogs.com/txq157/p/6111593.html
