function?tree?=?build_tree（train_features?train_targets?discrete_dim?layervarargin）????
????
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%調用C4.5決策樹算法建立決策樹
%training_features：訓練樣本的特征??
%training_targets：訓練樣本所屬類別??
%discrete_dim：各個維度的特征是否是連續特征，0指的是連續特征??
%layer:節點所屬樹的層數
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
????
if?nargin>5
????pruning=varargin{1};
else
????pruning=35;
end
????????
[fea?L]=?size（train_features）;??
ale=?unique（train_targets）;??
tree.feature_tosplit=?0;??
tree.location=inf;??%初始化分裂位置是inf??
????????
if?isempty（train_features）???
????return????
end????
???????????
if?（（pruning?>?L）?||?（L?==?1）?||（length（ale）?==?1））?%如果剩余訓練樣本太?。ㄐ∮趐runing），或只剩一個，或只剩一類標簽，退出????
????his=?hist（train_targets?length（ale））;??%統計樣本的標簽，分別屬于每個標簽的數目??
????[num?largest]=?max（his）;?
????tree.value=?[];????
????tree.location??=?[];????
????tree.child=?ale（largest）;
????return????
end????
?????????
for?i?=?1:length（ale）?%遍歷判別標簽的數目???
????Pnode（i）?=?length（find（train_targets?==?ale（i）））?/?L;?
end???

%計算當前節點的信息熵?
Inode?=?-sum（Pnode.*log2（Pnode））;????
????????
el=?zeros（1?fea）;??%記錄每個特征的信息增益率??
location=?ones（1?fea）*inf;
????????
for?i?=?1:fea?%遍歷每個特征????
????data=?train_features（i:）;?
????pe=?unique（data）;????
????nu=?length（pe）;???
????if?（discrete_dim（i））?%離散特征???
????????node=?zeros（length（ale）?nu）;
????????for?j?=?1:length（ale）?%遍歷每個標簽????
????????????for?k?=?1:nu?%遍歷每個特征值????
????????????????indices?????=?find（（train_targets?==?ale（j））?&&?（train_features（i:）?==?pe（k）））;????
????????????????node（jk）??=?length（indices）;
????????????end????
????????end????
????????rocle=?sum（node）;
????????P1=?repmat（rocle?length（ale）?1）;
????????P1=?P1?+?eps*（P1==0）;
????????node=?node./P1;
????????rocle=?rocle/sum（rocle）;
????????info=?sum（-node.*log（eps+node）/log（2））;??%每個特征分別計算信息熵eps是為了防止對數為1?
????????el（i）?=?（Inode-sum（rocle.*info））/（-sum（rocle.*log（eps+rocle）/log（2）））;?%信息增益率???
????else???%連續特征
????????node=?zeros（length（ale）?2）;
????????
????????[sorted_data?indices]?=?sort（data）;
????????sorted_targets?=?train_targets（indices）;
????????
????????%計算分裂信息度量??
?????????I?=?zeros（1nu）;??
?????????spl=?zeros（1?nu）;??
?????????for?j?=?1:nu-1??%特征i有Nbins個連續值，設定Nbins-1個可能的分割點，對每個分割點計算信息增益率
?????????????node（:?1）?=?hist（sorted_targets（find（sorted_data?<=?pe（j）））??ale）;??
?????????????node（:?2）?=?hist（sorted_targets（find（sorted_data?>?pe（j）））??ale）;???
?????????????Ps=?sum（node）/L;?
?????????????node=?node/L;?
?????????????rocle=?sum（node）;????
?????????????P1=?repmat（rocle?length（ale）?1）;?
?????????????P1=?P1?+?eps*（P1==0）;????
?????????????info=?sum（-node./P1.*log（eps+node./P1）/log（2））;?%信息增益
?????????????I（j）=?Inode?-?sum（info.*Ps）;???
?????????????spl（j）?=I（j）/（-sum（Ps.*log（eps+Ps）/log（2）））;??%第j個分割點的信息增益率
?????????end??
??
???????[~?s]?=?max（I）;??%求所有分割點的最大信息增益率
???????el（i）?=?spl（s）;??
???????location

?屬性????????????大小?????日期????時間???名稱
-----------?---------??----------?-----??----

?????文件????????353??2018-04-14?21:41??C4.5決策樹分類大作業\readme.txt

?????文件????2635336??2018-04-26?22:44??C4.5決策樹分類大作業\提交\C4.5決策樹分類大作業.docx

?????文件????????687??2018-04-14?21:19??C4.5決策樹分類大作業\提交\readme.txt

?????文件???????5342??2018-04-26?22:32??C4.5決策樹分類大作業\提交\實驗代碼\build_tree.m

?????文件???????1382??2018-04-13?19:08??C4.5決策樹分類大作業\提交\實驗代碼\C4_5.m

?????文件????????586??2018-04-13?19:14??C4.5決策樹分類大作業\提交\實驗代碼\cal_accuracy.m

?????文件????????737??2018-04-13?19:12??C4.5決策樹分類大作業\提交\實驗代碼\discreteOrContinue.m

?????文件???????3997??2018-04-14?21:50??C4.5決策樹分類大作業\提交\實驗代碼\magnify.m

?????文件???????1790??2018-04-13?19:08??C4.5決策樹分類大作業\提交\實驗代碼\predict.m

?????文件???????1000??2018-04-14?21:46??C4.5決策樹分類大作業\提交\實驗代碼\runtime_C4_5.mat

?????文件????????996??2018-04-14?21:46??C4.5決策樹分類大作業\提交\實驗代碼\runtime_CART.mat

?????文件???????1389??2018-04-26?22:44??C4.5決策樹分類大作業\提交\實驗代碼\tree.mat

?????文件???????1869??2018-04-26?22:36??C4.5決策樹分類大作業\提交\實驗代碼\wine.m

?????文件??????11746??2018-04-13?10:41??C4.5決策樹分類大作業\提交\實驗代碼\WineData.txt

?????文件?????248540??2018-04-13?19:47??C4.5決策樹分類大作業\資料\1.png

?????文件???????3268??2018-04-13?19:17??C4.5決策樹分類大作業\資料\C4.5.PNG

?????文件???????3298??2018-04-13?19:18??C4.5決策樹分類大作業\資料\C4.5_1.PNG

?????文件???????2187??2018-04-13?19:24??C4.5決策樹分類大作業\資料\C4.5_11.PNG

?????文件???????2330??2018-04-13?19:27??C4.5決策樹分類大作業\資料\C4.5_111.PNG

?????文件???????2315??2018-04-13?19:27??C4.5決策樹分類大作業\資料\C4.5_112.PNG

?????文件???????2347??2018-04-13?19:24??C4.5決策樹分類大作業\資料\C4.5_12.PNG

?????文件???????2338??2018-04-13?19:25??C4.5決策樹分類大作業\資料\C4.5_13.PNG

?????文件???????3275??2018-04-13?19:25??C4.5決策樹分類大作業\資料\C4.5_14.PNG

?????文件???????3230??2018-04-13?19:19??C4.5決策樹分類大作業\資料\C4.5_2.PNG

?????文件??????23313??2018-04-13?23:17??C4.5決策樹分類大作業\資料\C4.5分類準確度.fig

?????文件?????151884??2018-04-14?11:02??C4.5決策樹分類大作業\資料\C4.5和CART算法復雜度分析.fig

?????文件??????25050??2018-04-14?11:56??C4.5決策樹分類大作業\資料\C4.5和CART運行結果.fig

?????文件??????49658??2018-04-14?11:54??C4.5決策樹分類大作業\資料\C4.5和CART運行結果.png

?????文件????1914947??2018-04-14?11:53??C4.5決策樹分類大作業\資料\C4.5和CART運行結果_ps簡單處理.png

?????文件?????125166??2018-04-14?20:52??C4.5決策樹分類大作業\資料\C4.5數據分類大作業.docx

............此處省略21個文件信息