读书笔记：机器学习实战(2)——章3的决策树代码和个人理解与注释

首先是对于决策树的个人理解：
通过寻找最大信息增益（或最小信息熵）的分类特征，从部分已知类别的数据中提取分类规则的一种分类方法。
信息熵：

其中，log底数为2，额，好吧，图片我从百度截的。。
这里只解释到它是一种信息的期望值，深入的请看维基百科

http://zh.wikipedia.org/zh-sg/熵_(信息论)

信息增益：划分数据集前后的信息发生的变化（原书定义）
实际应用想要找到具有最大信息增益的分类树结构，就是使原始数据的信息熵减去分类后的信息熵的差值最大，原始数据的信息熵可以理解为常数，那么想要最大信息增益，也就是要寻找一种分类方法，是按照分类方法分类后的数据集的信息熵最小。（另：也可以选择“不纯度”或“错误率”作为评估参数，不纯度，维基百科下，错误率就是字面意思）
所以找到最优分类树结构代码如下：

def chooseBestFeatureToSplit(dataSet):numFeatures = len(dataSet[0]) - 1      #the last column is used for the labelsbaseEntropy = calcShannonEnt(dataSet)bestInfoGain = 0.0; bestFeature = -1for i in range(numFeatures):        #iterate over all the featuresfeatList = [example[i] for example in dataSet]#create a list of all the examples of this featureuniqueVals = set(featList)       #get a set of unique valuesnewEntropy = 0.0for value in uniqueVals:subDataSet = splitDataSet(dataSet, i, value)prob = len(subDataSet)/float(len(dataSet))newEntropy += prob * calcShannonEnt(subDataSet)     infoGain = baseEntropy - newEntropy     #calculate the info gain; ie reduction in entropyif (infoGain > bestInfoGain):       #compare this to the best gain so farbestInfoGain = infoGain         #if better than current best, set to bestbestFeature = ireturn bestFeature

按照“寻找最大信息增益的方式”，找到对于已知类别的一批数据（训练集）的最优决策树，然后用这个树结构去分类未知数据（测试集），整体代码如下：

#!/usr/bin/env python
# coding=utf-8
'''
Created on Oct 12, 2010
Decision Tree Source Code for Machine Learning in Action Ch. 3
@author: Peter Harrington
'''
from math import log
import operatordef createDataSet():dataSet = [[1, 1, 'yes'],[1, 1, 'yes'],[1, 0, 'no'],[0, 1, 'no'],[0, 1, 'no']]labels = ['no surfacing','flippers']#change to discrete valuesreturn dataSet, labelsdef calcShannonEnt(dataSet):# 计算香侬熵numEntries = len(dataSet)labelCounts = {}# 存储特征的字典for featVec in dataSet: #the the number of unique elements and their occurancecurrentLabel = featVec[-1]# 取最后一个元素，即该组特征的labelif currentLabel not in labelCounts.keys(): labelCounts[currentLabel] = 0# 如果没有，增加新key，value初始化为0labelCounts[currentLabel] += 1# 对应key的值累计shannonEnt = 0.0for key in labelCounts:prob = float(labelCounts[key])/numEntriesshannonEnt -= prob * log(prob,2) #log base 2# shannon公式：shanonEnt ＝（负的）求和（i.prob*log(i.prob,2))return shannonEntdef splitDataSet(dataSet, axis, value):retDataSet = []for featVec in dataSet:if featVec[axis] == value:reducedFeatVec = featVec[:axis]     #chop out axis used for splittingreducedFeatVec.extend(featVec[axis+1:])# 简单的分片，除去分类特征，余下的添加到容器中retDataSet.append(reducedFeatVec)return retDataSetdef chooseBestFeatureToSplit(dataSet):numFeatures = len(dataSet[0]) - 1      #the last column is used for the labelsbaseEntropy = calcShannonEnt(dataSet)bestInfoGain = 0.0; bestFeature = -1for i in range(numFeatures):        #iterate over all the featuresfeatList = [example[i] for example in dataSet]#create a list of all the examples of this featureuniqueVals = set(featList)       #get a set of unique valuesnewEntropy = 0.0for value in uniqueVals:subDataSet = splitDataSet(dataSet, i, value)prob = len(subDataSet)/float(len(dataSet))newEntropy += prob * calcShannonEnt(subDataSet)     infoGain = baseEntropy - newEntropy     #calculate the info gain; ie reduction in entropyif (infoGain > bestInfoGain):       #compare this to the best gain so farbestInfoGain = infoGain         #if better than current best, set to bestbestFeature = ireturn bestFeature                      #returns an integerdef majorityCnt(classList):classCount={}for vote in classList:if vote not in classCount.keys(): classCount[vote] = 0classCount[vote] += 1sortedClassCount = sorted(classCount.iteritems(), key=operator.itemgetter(1), reverse=True)return sortedClassCount[0][0]def createTree(dataSet,labels):classList = [example[-1] for example in dataSet]if classList.count(classList[0]) == len(classList): return classList[0]#stop splitting when all of the classes are equalif len(dataSet[0]) == 1: #stop splitting when there are no more features in dataSetreturn majorityCnt(classList)bestFeat = chooseBestFeatureToSplit(dataSet)bestFeatLabel = labels[bestFeat]myTree = {bestFeatLabel:{}}del(labels[bestFeat])featValues = [example[bestFeat] for example in dataSet]uniqueVals = set(featValues)for value in uniqueVals:subLabels = labels[:]       #copy all of labels, so trees don't mess up existing labelsmyTree[bestFeatLabel][value] = createTree(splitDataSet(dataSet, bestFeat, value),subLabels)return myTree                            def classify(inputTree,featLabels,testVec):firstStr = inputTree.keys()[0]secondDict = inputTree[firstStr]featIndex = featLabels.index(firstStr)key = testVec[featIndex]valueOfFeat = secondDict[key]if isinstance(valueOfFeat, dict): classLabel = classify(valueOfFeat, featLabels, testVec)else: classLabel = valueOfFeatreturn classLabeldef storeTree(inputTree,filename):import picklefw = open(filename,'w')pickle.dump(inputTree,fw)fw.close()def grabTree(filename):import picklefr = open(filename)return pickle.load(fr)if __name__ == '__main__':(dataSet, labels) = createDataSet();print dataSet;print labelsshannonEnt = calcShannonEnt(dataSet)print shannonEntmyTree = createTree(dataSet,labels )print 'mytree:'print myTree(dataSet, labels) = createDataSet();print classify(myTree,labels, [1,1])import treePlotter# treePlotter.createPlot(myTree)fr = open('lenses.txt')lenses = [inst.strip().split('\t') for inst in fr.readlines()]lensesLabels = ['age','prescript','astigmatic','tearRate']lensesTree = createTree(lenses,lensesLabels)print lensesTreetreePlotter.createPlot(lensesTree)

部分地方加入了中文注释，原著的那几行英文注释很好就没有再换成中文的。
之前没有详细看过决策树，以为它就是把分类逻辑变为树结构，多个if else，说说个人学习后，对于决策树的理解：
1.还是觉得它就是多个if else，树结构也可以这么理解吧
2.构建的过程或者说收敛条件是：最大信息增益（最小信息熵）
3.优点：可读性强，逻辑简单易懂，计算步骤不超过树的深度；缺点：极易过拟合，得到的树结构泛性不强
4.正因为过度追求最优解，导致决策树往往会过拟合，原著是通过构建以后的合并细小或相近分支，也就是“后置裁剪”，但是这样时间上有浪费，代表的是K-Fold Cross Validation，不断裁剪，评估当前的错误率，有点类似于整体求解后，再反过来找恰当的“early stop”；另一种就是著名的随机森林，系统复杂了，效果确实会好，额，随机森林具体的以后深度学习下补上。
下面是原著利用matplolib画图的代码：

'''
Created on Oct 14, 2010@author: Peter Harrington
'''
import matplotlib.pyplot as pltdecisionNode = dict(boxstyle="sawtooth", fc="0.8")
leafNode = dict(boxstyle="round4", fc="0.8")
arrow_args = dict(arrowstyle="<-")def getNumLeafs(myTree):numLeafs = 0firstStr = myTree.keys()[0]secondDict = myTree[firstStr]for key in secondDict.keys():if type(secondDict[key]).__name__=='dict':#test to see if the nodes are dictonaires, if not they are leaf nodesnumLeafs += getNumLeafs(secondDict[key])else:   numLeafs +=1return numLeafsdef getTreeDepth(myTree):maxDepth = 0firstStr = myTree.keys()[0]secondDict = myTree[firstStr]for key in secondDict.keys():if type(secondDict[key]).__name__=='dict':#test to see if the nodes are dictonaires, if not they are leaf nodesthisDepth = 1 + getTreeDepth(secondDict[key])else:   thisDepth = 1if thisDepth > maxDepth: maxDepth = thisDepthreturn maxDepthdef plotNode(nodeTxt, centerPt, parentPt, nodeType):createPlot.ax1.annotate(nodeTxt, xy=parentPt,  xycoords='axes fraction',xytext=centerPt, textcoords='axes fraction',va="center", ha="center", bbox=nodeType, arrowprops=arrow_args )def plotMidText(cntrPt, parentPt, txtString):xMid = (parentPt[0]-cntrPt[0])/2.0 + cntrPt[0]yMid = (parentPt[1]-cntrPt[1])/2.0 + cntrPt[1]createPlot.ax1.text(xMid, yMid, txtString, va="center", ha="center", rotation=30)def plotTree(myTree, parentPt, nodeTxt):#if the first key tells you what feat was split onnumLeafs = getNumLeafs(myTree)  #this determines the x width of this treedepth = getTreeDepth(myTree)firstStr = myTree.keys()[0]     #the text label for this node should be thiscntrPt = (plotTree.xOff + (1.0 + float(numLeafs))/2.0/plotTree.totalW, plotTree.yOff)plotMidText(cntrPt, parentPt, nodeTxt)plotNode(firstStr, cntrPt, parentPt, decisionNode)secondDict = myTree[firstStr]plotTree.yOff = plotTree.yOff - 1.0/plotTree.totalDfor key in secondDict.keys():if type(secondDict[key]).__name__=='dict':#test to see if the nodes are dictonaires, if not they are leaf nodes   plotTree(secondDict[key],cntrPt,str(key))        #recursionelse:   #it's a leaf node print the leaf nodeplotTree.xOff = plotTree.xOff + 1.0/plotTree.totalWplotNode(secondDict[key], (plotTree.xOff, plotTree.yOff), cntrPt, leafNode)plotMidText((plotTree.xOff, plotTree.yOff), cntrPt, str(key))plotTree.yOff = plotTree.yOff + 1.0/plotTree.totalD
#if you do get a dictonary you know it's a tree, and the first element will be another dictdef createPlot(inTree):fig = plt.figure(1, facecolor='white')fig.clf()axprops = dict(xticks=[], yticks=[])createPlot.ax1 = plt.subplot(111, frameon=False, **axprops)    #no ticks#createPlot.ax1 = plt.subplot(111, frameon=False) #ticks for demo puropses plotTree.totalW = float(getNumLeafs(inTree))plotTree.totalD = float(getTreeDepth(inTree))plotTree.xOff = -0.5/plotTree.totalW; plotTree.yOff = 1.0;plotTree(inTree, (0.5,1.0), '')plt.show()#def createPlot():
#    fig = plt.figure(1, facecolor='white')
#    fig.clf()
#    createPlot.ax1 = plt.subplot(111, frameon=False) #ticks for demo puropses
#    plotNode('a decision node', (0.5, 0.1), (0.1, 0.5), decisionNode)
#    plotNode('a leaf node', (0.8, 0.1), (0.3, 0.8), leafNode)
#    plt.show()def retrieveTree(i):listOfTrees =[{'no surfacing': {0: 'no', 1: {'flippers': {0: 'no', 1: 'yes'}}}},{'no surfacing': {0: 'no', 1: {'flippers': {0: {'head': {0: 'no', 1: 'yes'}}, 1: 'no'}}}}]return listOfTrees[i]#createPlot(thisTree)

读书笔记：机器学习实战(2)——章3的决策树代码和个人理解与注释相关推荐

强化学习（RLAI）读书笔记第十六章Applications and Case Studies（不含alphago）
强化学习(RLAI)读书笔记第十六章Applications and Case Studies(不含alphago) 16.1 TD-Gammon 16.2 Samuel's Checkers Pla ...
《Android开发艺术探索》读书笔记 (3) 第3章 View的事件体系
本节和<Android群英传>中的第五章Scroll分析有关系,建议先阅读该章的总结第3章 View的事件体系 3.1 View基本知识 (1)view的层次结构:ViewGroup也是 ...
Think in Java第四版读书笔记10 第16章数组
Think in Java第四版读书笔记10 第16章数组数组和容器很像但他们有一些差别 16.1 数组为什么特殊数组与容器的区别主要在效率和存储类型效率:数组是简单的线性序列使得数组的 ...
Think in Java第四版读书笔记9第15章泛型
Think in Java第四版读书笔记9第15章泛型泛型:适用于很多很多的类型与其他语言相比 Java的泛型可能有许多局限但是它还是有很多优点的. 本章介绍java泛型的局限和优势以及ja ...
《统计学习方法》读书笔记——机器学习常用评价指标
传送门 <统计学习方法>读书笔记--机器学习常用评价指标 <统计学习方法>读书笔记--感知机(原理+代码实现) <统计学习方法>读书笔记--K近邻法(原理+代码实现 ...
【Python自然语言处理】读书笔记：第五章：分类和标注词汇
jupyter 版请见我的github:https://github.com/JackKuo666/Python_nlp_notes [Python自然语言处理]读书笔记:第五章:分类和标注词汇本章 ...
马丁福勒《UML精粹》读书笔记_第四章
第四章顺序图顺序图是一个use case的一种实现.当考察单个use case内部若干对象的行为时,就应使用顺序图. 可参考"高焕堂<嵌入式UML设计>读书笔记_第五章&qu ...
正义之心读书笔记：第7章自由和保守主义的本质——5大道德基础
人不是经济人经济人:在做人生决定时对各种选择都作了充分的考虑,且仅有一个影响因素:个人私利.人无论做任何事都要用最低的成本换取最丰厚的回报. 实际上,人有一系列的道德基础.在作者看来,主要有5类:关 ...
黑帽python第二版（Black Hat Python 2nd Edition）读书笔记之第四章使用SCAPY掌控网络（2）Scapy实现ARP缓存投毒
黑帽python第二版(Black Hat Python 2nd Edition)读书笔记之第四章使用SCAPY掌控网络(2)Scapy实现ARP缓存投毒文章目录黑帽python第二版(Bl ...

读书笔记：机器学习实战(2)——章3的决策树代码和个人理解与注释

读书笔记：机器学习实战(2)——章3的决策树代码和个人理解与注释相关推荐

最新文章

热门文章