西電數(shù)據(jù)挖掘決策樹(shù)算法

1、數(shù)據(jù)挖掘算法實(shí)驗(yàn)報(bào)告1) 實(shí)驗(yàn)題目基于決策樹(shù)的分類(lèi)算法，屬性的選擇采用ID3或C4.5策略，采用如下的數(shù) 據(jù)建立分類(lèi)決策樹(shù)。ageincomestudentcredit ratingbuys computer=30highnofairno40mediumnofairves40lowyesrairyes4Qlowvesexcellentno31. .40lowvesexcellentves=30mediumnofairno40mediumyesfairyes40mediumnoexcelentno2)算法基本思想的描述ID3選擇具有最高信息熵增益的屬性作為分裂屬性，基于這種原則我們 首先可以算出

2、初始集合的熵，然后分別求出以各個(gè)屬性為分裂屬性時(shí)的熵， 然后將通過(guò)上面得到的數(shù)據(jù)算出以各個(gè)屬性為分裂屬性時(shí)的信心增益，選擇 具有最大的信息增益屬性作為我們的分裂屬性。3) 編程實(shí)現(xiàn)算法#include #include #include using namespace std;#define SIZE 14struct Data(char age10;char income10;char student10;char credit_rating20;char buys_computer10;；Data data SIZE=(=30，high，氣o，fair，氣o,40,medium,no,fa

3、ir,yes,40,low,yes,fair,yes,40,low,yes,excellent,no,31.40,low,yes,excellent,yes,=30,medium,no,fair,no,40,medium,yes,fair,yes,40,medium,no,excellent,no；double calculate(double a,double b);void origin_entropy(Data data,double &entropy);void age_entropy(Data data,double &entropy);void income_entropy(Dat

4、a data,double &entropy);void student_entropy(Data data,double &entropy);void credit_rating_entropy(Data data,double &entropy);int main()(double origin=0,age=0,student=0,credit_rating=0,income=0;origin_entropy(data,origin);age_entropy(data,age);student_entropy(data,student);income_entropy(data,income

5、);credit_rating_entropy(data,credit_rating);coutinfo(D)=originendl;cout用 age 作為分裂屬性時(shí)：偵熵 info(age)(D)=agett信息增益為：origin-ageendl;cout用 income 作為分裂屬性時(shí)：n熵 info(income)(D)=incomett 信 息 增 益 為：origin-incomeendl;cout用 student 作為分裂屬性時(shí)：n熵 info(student)(D)=):studenttt 信 息 增 益 為：origin-studentendl;cout用 credit_

6、rating 作為分裂屬性時(shí)：n 熵 info(credit_rating)(D)=credit_ratingt 信息增益 為：origin-credit_ratingendl;return 0;double calculate(double a,double b)(if(a=0)return 0;elsereturn (a/b)*log10(b/a)/log10(2);void origin_entropy(Data data,double &entropy)(int i;double yes=0, no=0;for( i=0;iSIZE;i+)(if(strcmp(datai.buys_c

7、omputer,yes)=0)yes+;elseno+;entropy二calculate(yes,SIZE)+calculate(no,SIZE);void age_entropy(Data data,double &entropy)(double youth3=0;double middle3=0;double old3 = 0;for (int i=0;iSIZE;i+)(if(strcmp(datai.age,=30)=0)(strcmp(datai.buys_computer,yes)=0)?youth0+:youth1+ ;youth2+;else if(strcmp(datai.

8、age,31.40)=0)(strcmp(datai.buys_computer,yes)=0)?middle0+:middle1 +;middle2+;else(strcmp(datai.buys_computer,yes)=0)?old0+:old1+;old 2+;entropy二youth2/SIZE*(calculate(youth0,youth2)+calculate( youth1,youth2)+middle2/SIZE*(calculate(middle0,middle2 )+calculate(middle1,middle2)+old2/SIZE*(calculate(ol

9、d0 ,old2)+calculate(old1,old2);void income_entropy(Data data,double &entropy)(double high3 = 0;double medium3=0;double low3 = 0;for (int i=0;iSIZE;i+)(if(strcmp(datai.income,high)=0)(strcmp(datai.buys_computer,yes)=0)?high0+:high1+;h igh2+;else if(strcmp(datai.income,medium)=0)(strcmp(datai.buys_com

10、puter,yes)=0)?medium0+:medium1 +;medium2+;else(strcmp(datai.buys_computer,yes)=0)?low0+:low1+;low 2+;entropy=high2/SIZE*(calculate(high0,high2)+calculate(hig h1,high2)+medium2/SIZE*(calculate(medium0,medium2)+ca lculate(medium1,medium2)+low2/SIZE*(calculate(low0,low 2)+calculate(low1,low2);void stud

11、ent_entropy(Data data,double &entropy)(double yes3 = 0;double no3 = 0;for (int i=0;iSIZE;i+)(if(strcmp(datai.student,yes)(strcmp(datai.buys_computer,yes)？yes0+:yes1+;yes2+7else(strcmp(datai.buys_computer,yes)？no0+:no1+;no2+;entropy=yes2/SIZE*(calculate(yes0,yes2)+calculate(yes1,yes2)+no2/SIZE*(calculate(no0,no2)+calculate(no1,no 2);void credit_rating_entropy(Data data,double &entropy)(double fair3 = 0;double excellent3=0;for (int i=0;iSIZE;i+)(if(strcmp(datai.credit_rating,fair)(strcmp(datai.buys_computer,yes)?fair0+:fair1+;fair2+;else(strcmp(datai.buys_computer,yes)?excellent0+:exce