GeneralizedLinearAlgorithm

SparkMllib涉及到的算法

  • Classification
  • Linear Support Vector Machines (SVMs)
  • Logistic regression
  • Regression
  • Linear least squares, Lasso, and ridge regression
  • Streaming linear regression

GeneralizedLinearAlgorithm

GLA,通用线性算法,作为通用回归算法(regression)和分类算法(classification)的抽象算法,run函数中实现了算法的流程,并最终产生通用线性模型。抽象算法流程主要包括:addIntercept, useFeatureScaling, 调用optimizer计算参数等。其中optimizer是抽象类,具体的线性算法需要指定具体的optimizer实现类。【模板模式——子类的训练大多调用了GLA中的run函数】

进一步地,optimizer是算法的核心,也是迭代算法的逻辑所在。具体类包括梯度下降等算法。这些优化算法又可以抽象并归纳出两个模块:梯度计算器(Gradient)和参数更新(Updater)。

综上,对于某一个具体的线性算法时主要一个优化器,而优化器是通过选择梯度计算器和参数更新器的组合来得到。【bridge pattern】

当前版本中GLA有多个子类,覆盖了多个分类和回归算法:

  • LassoWithSGD (org.apache.spark.mllib.regression)
  • LinearRegressionWithSGD (org.apache.spark.mllib.regression)
  • RidgeRegressionWithSGD (org.apache.spark.mllib.regression)
  • LogisticRegressionWithLBFGS (org.apache.spark.mllib.classification)
  • LogisticRegressionWithSGD (org.apache.spark.mllib.classification)
  • SVMWithSGD (org.apache.spark.mllib.classification)

例如,LinearRegressionWithSGD算法,对应的OptimizerGradientDecent,对应的梯度计算器是LeaseSquareGradentx位置的梯度$$\vec{\triangledown} = (\vec{x}^T\cdot\vec{w} - y)\cdot\vec{x}$$,对应的更新器为SimpleUpdater,更新算法是$$\vec{w} =: \vec{w} - \alpha / \sqrt{iter}\cdot\vec{\triangledown}/m$$

GeneralizedLinearModel

Algorithm will create a model by calling:

def createModel(...): M

**GeneralizedLinearModel **(GLM) represents a model trained using GeneralizedLinearAlgorithm. GLMs consist of a weight vector and an intercept.

Parameters:

  • weights - Weights computed for every feature.
  • intercept - Intercept computed for this model.

    基本上每个线性算法都对应到一个线性模型:
  • LassoModel
  • LinearRegressionModel
  • LogisticRegressionModel
  • RidgeRegressionModel
  • SVMModel

GradientDecent梯度下降法

梯度下降法对convex函数必然能求解。

function [theta, J_history] = gradientDescent(X, y, theta, alpha, num_iters)
%GRADIENTDESCENT Performs gradient descent to learn theta
% theta = GRADIENTDESCENT(X, y, theta, alpha, num_iters) updates theta by
% taking num_iters gradient steps with learning rate alpha % Initialize some useful values
m = length(y); % number of training examples
J_history = zeros(num_iters, 1);
for iter = 1:num_iters
% ====================== YOUR CODE HERE ======================
% Instructions: Perform a single gradient step on the parameter vector
% theta.
%
% Hint: While debugging, it can be useful to print out the values
% of the cost function (computeCost) and gradient here.
%
theta = theta - alpha * X' * (X * theta - y) / m;
% ============================================================
% Save the cost J in every iteration
J_history(iter) = computeCost(X, y, theta); end

代码中可以发现多处使用了梯度下降法,包括ann,LR/SVM分类算法,Lasso/Linear/Ridge回归等算法。

主要构造函数:

class GradientDescent private[spark] (
private var gradient: Gradient,
private var updater: Updater)
extends Optimizer with Logging {
private var stepSize: Double = 1.0
private var numIterations: Int = 100
private var regParam: Double = 0.0
private var miniBatchFraction: Double = 1.0
private var convergenceTol: Double = 0.001

其中gradient为计算梯度的公式,updater为根据梯度的值去更新权重的公式。除了这两个参数之外,还有算法训练过程中的一些技术参数。

算法的基本参数设定之后便,再给定训练样本,便可以训练得到模型权重:

def optimize(
data: RDD[(Double, Vector)],
initialWeights: Vector): Vector = {
val (weights, _) = GradientDescent.runMiniBatchSGD(
data,
gradient,
updater,
stepSize,
numIterations,
regParam,
miniBatchFraction,
initialWeights,
convergenceTol)
weights
}

=> 调用迭代梯度下降法

def runMiniBatchSGD(data: RDD[(Double, Vector)],
gradient: Gradient,
updater: Updater,
stepSize: Double,
numIterations: Int,
regParam: Double,
miniBatchFraction: Double,
initialWeights: Vector,
convergenceTol: Double): (Vector, Array[Double])

Run stochastic gradient descent (SGD) in parallel using mini batches. In each iteration, we sample a subset (fraction miniBatchFraction) of the total data in order to compute a gradient estimate. Sampling, and averaging the subgradients over this subset is performed using one standard spark map-reduce in each iteration.

Parameters:

  • data - Input data for SGD. RDD of the set of data examples, each of the form (label, [feature values]).
  • gradient - Gradient object (used to compute the gradient of the loss function of one single data example)
  • updater - Updater function to actually perform a gradient step in a given direction.
  • stepSize - initial step size for the first step
  • numIterations - number of iterations that SGD should be run.
  • regParam - regularization parameter
  • miniBatchFraction - fraction of the input data set that should be used for one iteration of SGD. Default value 1.0.
  • initialWeights - initial weights for model training.
  • convergenceTol - Minibatch iteration will end before numIterations if the relative difference between the current weight and the previous weight is less than this value. In measuring convergence, L2 norm is calculated. Default value 0.001. Must be between 0.0 and 1.0 inclusively.

Returns:

A tuple containing two elements. The first element is a column matrix containing weights for every feature, and the second element is an array containing the stochastic loss computed for every iteration.

迭代梯度下降法 (SGD)

上述分析可以发现梯度下降法采用的实现法法是SGD,该算法对模型进行迭代更新,非常适合分布式计算,算法流程:

  1. Sample a subset of data or just use all the data;
  2. [MAPPER] for each entry, calculate gradient and model loss by Gradient $$(g^{(i)}, l^{(i)}) := \cal{Gradient}(x^{(i)}, y^{(i)}, \theta)$$;
  3. [AGGREGATION] sum of gradient = $$\sum_ig^{(i)}$$, sum of loss = $$\sum_i l^{(i)}$$, counter = $$\sum_i 1$$;
  4. Updater $$\theta' := \cal{Update}(\theta, \sum_ig^{(i)} / \sum_i 1, \alpha, ...)$$;
  5. if $$||\theta'-\theta||_2 / ||\theta'||_2 < convergenceTol$$ finish, else $$\theta := \theta'$$, repeat from step 1.

各线性算法的G/U组合

Algorithm optimizer Gradient updater
LinearRegressionWithSGD GradientDescent LeastSquares Simple
LassoWithSGD GradientDescent LeastSquares L1Updater
RidgeRegressionWithSGD GradientDescent LeastSquares SquaredL2
LogisticRegressionWithSGD GradientDescent Logistic SquaredL2
LogisticRegressionWithLBFGS LBFGS Logistic SquaredL2
SVMWithSGD GradientDescent Hinge SquaredL2

GradientDecent以及SGD算法的流程在上面已经描述,接下来具体分析各种梯度计算和权重更新算法:

LeaseSquareGradent

lost function:

\[L = 1/n ||X \cdot w - y||^2
\]

\[\vec x_i$$位置的梯度$$\vec{\triangledown_i} = (\vec{x_i}^T\cdot\vec{w} - \vec{y}_i)\cdot\vec{x_i}
\]

SimpleUpdater

A simple updater for gradient descent without any regularization.

更新算法是$$\vec{w} := \vec{w} - \alpha / \sqrt{iter}\cdot\vec{\triangledown}$$,

其中$$ \vec{\triangledown} = \sum_i{\vec{\triangledown_i} / \sum_i {i}}$$

算法迭代过程中整体的步长会随着迭代次数而逐渐减小,

SquaredL2Updater

L2 regularized problems:

\[R(w) = \lambda /2 ||w||^2\\
L = 1/n ||X \cdot w - y||^2 + R(w),\\
\text{then:} \frac{dL}{dw}=\vec{\triangledown} + \lambda w\\
w:=w' - \alpha \cdot \frac{dL}{dw}\\
\ \ \ = (1-\alpha \lambda)\vec{w} - \alpha / \sqrt{iter}\cdot\vec{\triangledown}\]

LinearRegressionWithSGD比较,RidgeRegressionWithSGD的优化器依然是LeastSquaresGradient,训练样本计算Gradient额算法也相同,不通点是在给定Gradient结果基础上更新权重时,old权重需要额外乘以一个系数。

GeneralizedLinearAlgorithm in Spark MLLib的更多相关文章

  1. Spark MLlib之线性回归源代码分析

    1.理论基础 线性回归(Linear Regression)问题属于监督学习(Supervised Learning)范畴,又称分类(Classification)或归纳学习(Inductive Le ...

  2. Spark MLlib - LFW

    val path = "/usr/data/lfw-a/*" val rdd = sc.wholeTextFiles(path) val first = rdd.first pri ...

  3. 《Spark MLlib机器学习实践》内容简介、目录

      http://product.dangdang.com/23829918.html Spark作为新兴的.应用范围最为广泛的大数据处理开源框架引起了广泛的关注,它吸引了大量程序设计和开发人员进行相 ...

  4. Spark MLlib 之 Basic Statistics

    Spark MLlib提供了一些基本的统计学的算法,下面主要说明一下: 1.Summary statistics 对于RDD[Vector]类型,Spark MLlib提供了colStats的统计方法 ...

  5. Spark MLlib Data Type

    MLlib 支持存放在单机上的本地向量和矩阵,也支持通过多个RDD实现的分布式矩阵.因此MLlib的数据类型主要分为两大类:一个是本地单机向量:另一个是分布式矩阵.下面分别介绍一下这两大类都有哪些类型 ...

  6. Spark MLlib - Decision Tree源码分析

    http://spark.apache.org/docs/latest/mllib-decision-tree.html 以决策树作为开始,因为简单,而且也比较容易用到,当前的boosting或ran ...

  7. Spark入门实战系列--8.Spark MLlib(上)--机器学习及SparkMLlib简介

    [注]该系列文章以及使用到安装包/测试数据 可以在<倾情大奉送--Spark入门实战系列>获取 .机器学习概念 1.1 机器学习的定义 在维基百科上对机器学习提出以下几种定义: l“机器学 ...

  8. Spark入门实战系列--8.Spark MLlib(下)--机器学习库SparkMLlib实战

    [注]该系列文章以及使用到安装包/测试数据 可以在<倾情大奉送--Spark入门实战系列>获取 .MLlib实例 1.1 聚类实例 1.1.1 算法说明 聚类(Cluster analys ...

  9. spark mllib配置pom.xml错误 Multiple markers at this line Could not transfer artifact net.sf.opencsv:opencsv:jar:2.3 from/to central (https://repo.maven.apache.org/maven2): repo.maven.apache.org

    刚刚spark mllib,在maven repository网站http://mvnrepository.com/中查询mllib后得到相关库的最新dependence为: <dependen ...

随机推荐

  1. [SoapUI] Groovy获取HTTP Status

    def value = messageExchange.responseHeaders["#status#"] def httpResponseHeaders = context. ...

  2. ajax异步请求该嵌套还是并列?

    因为要查询两个数据库表才能确定我所需要的数据范围,所以前台js得发两次ajax请求.问题就是,这两个请求是嵌套着写:{发,接{发,接}}:还是并列着写:{发,接},{发,接} 好? 答案:如果2次aj ...

  3. Linux搭建SVN

    Linux搭建SVN 服务器 1 安装SVN 官网下载:http://subversion.apache.org/packages.html SVN客户端:TortoiseSVN,官网下载:http: ...

  4. [ES]ES查询指南

    我们通常用用_cat API检测集群是否健康. 确保9200端口号可用: curl 'localhost:9200/_cat/health?v' 绿色表示一切正常, 黄色表示所有的数据可用但是部分副本 ...

  5. NETSHARP的JAVA开发环境配置

    一:JAVA配置 1. netsharp使用java1.8/1.7版本,本文使用1.8版本 2.jdk下载地址:http://www.oracle.com/technetwork/java/javas ...

  6. Java虚拟机的相关笔记

    1.垃圾GC回收事件Minor GC(只清除新生代),Full GC(清除新生代和老年代),Major GC(清除新生.老年代和持久代). 2.堆分为新生代.老年代和持久代,持久代一般存放静态文件. ...

  7. Web 开发

    Django(发音:[`dʒæŋɡəʊ]) 是一个开放源代码的Web应用框架,由Python写成.采用了MTV的框架模式,模型(Model).模板(Template)和视图(Views).

  8. 2019.01.23 hdu3377 Plan(轮廓线dp)

    传送门 题意简述:给一个n*m的带权矩阵,求从左上角走到右下角的最大分数,每个格子只能经过最多一次,n,m≤9n,m\le9n,m≤9. 思路: 考虑轮廓线dpdpdp,但这道题并没有出现回路的限制因 ...

  9. JSON.parse和JSON.stringify的区别

    JSON.stringify()的作用是将 JavaScript 值转换为 JSON 字符串, 而JSON.parse()可以将JSON字符串转为一个对象. 简单点说,它们的作用是相对的,我用JSON ...

  10. Spring-WebSocket

    WebSocket Sockjs Stoup (消息订阅发布) 添加依赖 <!-- 添加依赖 --> <dependencies> <dependency> < ...