matlab鑷甫鍚勭鍒嗙被鍣ㄧ殑浣跨敤绀轰緥

Posted 2020-12-27

鏍囩锛?a href='http://www.mamicode.com/so/1/%e5%8f%82%e8%80%83' title='鍙傝€?>鍙傝€?/a>   ati   tps   绀轰緥   dom   forest   ada   鏂规硶鎬荤粨   struct   

鍏ㄦ枃杞嚦 u014114990鐨勪笓鏍?/a>锛宎ddress锛歨ttps://blog.csdn.net/u014114990/article/details/51067059

涓汉淇敼寤鸿锛?/p>

璁粌鏁版嵁鍜岄獙璇佹暟鎹洿鎺ヤ互M*N琛ㄧず锛堣琛ㄧず鐗瑰緛涓暟锛屽垪涓烘牱鏈釜鏁帮級锛岃繖鏍峰彲浠ラ伩鍏嶅悗缁殑杞疆鎿嶄綔銆?/p>

 

鐩墠浜嗚В鍒扮殑MATLAB涓垎绫诲櫒鏈夛細K杩戦偦鍒嗙被鍣紝闅忔満妫灄鍒嗙被鍣紝鏈寸礌璐濆彾鏂紝闆嗘垚瀛︿範鏂规硶锛岄壌鍒垎鏋愬垎绫诲櫒锛屾敮鎸佸悜閲忔満銆傜幇灏嗗叾涓昏鍑芥暟浣跨敤鏂规硶鎬荤粨濡備笅锛屾洿澶氱粏鑺傞渶鍙傝€僊ATLAB 甯姪鏂囦欢銆?br>璁?br>銆€銆€璁粌鏍锋湰锛歵rain_data             % 鐭╅樀锛屾瘡琛屼竴涓牱鏈紝姣忓垪涓€涓壒寰?br>銆€銆€璁粌鏍锋湰鏍囩锛歵rain_label       % 鍒楀悜閲?br>銆€銆€娴嬭瘯鏍锋湰锛歵est_data
銆€銆€娴嬭瘯鏍锋湰鏍囩锛歵est_label
 
K杩戦偦鍒嗙被鍣?锛圞NN锛?br>mdl = ClassificationKNN.fit(train_data,train_label,鈥楴umNeighbors鈥?1);
predict_label   =       predict(mdl, test_data);
accuracy         =       length(find(predict_label == test_label))/length(test_label)*100
              
 
闅忔満妫灄鍒嗙被鍣紙Random Forest锛?br>B = TreeBagger(nTree,train_data,train_label);
predict_label = predict(B,test_data);
 
 
鏈寸礌璐濆彾鏂?锛圢a?ve Bayes锛?br>nb = NaiveBayes.fit(train_data, train_label);
predict_label   =       predict(nb, test_data);
accuracy         =       length(find(predict_label == test_label))/length(test_label)*100;
 
 
闆嗘垚瀛︿範鏂规硶锛圗nsembles for Boosting, Bagging, or Random Subspace锛?br>ens = fitensemble(train_data,train_label,鈥楢daBoostM1鈥?,100,鈥榯ree鈥?鈥榯ype鈥?鈥榗lassification鈥?;
predict_label   =       predict(ens, test_data);
 
 
閴村埆鍒嗘瀽鍒嗙被鍣紙discriminant analysis classifier锛?br>obj = ClassificationDiscriminant.fit(train_data, train_label);
predict_label   =       predict(obj, test_data);
 
 
鏀寔鍚戦噺鏈猴紙Support Vector Machine, SVM锛?br>SVMStruct = svmtrain(train_data, train_label);
predict_label  = svmclassify(SVMStruct, test_data)

浠ｇ爜锛?br>clc
clear all
 load(鈥榳dtFeature鈥?;
 
%  銆€銆€璁粌鏍锋湰锛歵rain_data             % 鐭╅樀锛屾瘡琛屼竴涓牱鏈紝姣忓垪涓€涓壒寰?br>% 銆€銆€璁粌鏍锋湰鏍囩锛歵rain_label       % 鍒楀悜閲?br>% 銆€銆€娴嬭瘯鏍锋湰锛歵est_data
% 銆€銆€娴嬭瘯鏍锋湰鏍囩锛歵est_label
 train_data = traindata鈥?br> train_label = trainlabel鈥?br> test_data = testdata鈥?br> test_label = testlabel鈥?br>%  K杩戦偦鍒嗙被鍣?锛圞NN锛?br>% mdl = ClassificationKNN.fit(train_data,train_label,鈥楴umNeighbors鈥?1);
% predict_label   =       predict(mdl, test_data);
% accuracy         =       length(find(predict_label == test_label))/length(test_label)*100
%               
%  94%
% 闅忔満妫灄鍒嗙被鍣紙Random Forest锛?br>% nTree = 5
% B = TreeBagger(nTree,train_data,train_label);
% predict_label = predict(B,test_data);
% 
% m=0;
% n=0;
% for i=1:50
%     if predict_label{i,1}>0
%         m=m+1;
%     end
%     if predict_label{i+50,1}<0
%         n=n+1;
%     end
% end
%
% s=m+n
% r=s/100
 
%  result 50%
 
% **********************************************************************
% 鏈寸礌璐濆彾鏂?锛圢a?ve Bayes锛?br>% nb = NaiveBayes.fit(train_data, train_label);
% predict_label   =       predict(nb, test_data);
% accuracy         =       length(find(predict_label == test_label))/length(test_label)*100;
%
%
% % 缁撴灉 81%
% % **********************************************************************
% % 闆嗘垚瀛︿範鏂规硶锛圗nsembles for Boosting, Bagging, or Random Subspace锛?br>% ens = fitensemble(train_data,train_label,鈥楢daBoostM1鈥?,100,鈥榯ree鈥?鈥榯ype鈥?鈥榗lassification鈥?;
% predict_label   =       predict(ens, test_data);
%
% m=0;
% n=0;
% for i=1:50
%     if predict_label(i,1)>0
%         m=m+1;
%     end
%     if predict_label(i+50,1)<0
%         n=n+1;
%     end
% end
%
% s=m+n
% r=s/100
 
% 缁撴灉 97%
% **********************************************************************
% 閴村埆鍒嗘瀽鍒嗙被鍣紙discriminant analysis classifier锛?br>% obj = ClassificationDiscriminant.fit(train_data, train_label);
% predict_label   =       predict(obj, test_data);
% 
% m=0;
% n=0;
% for i=1:50
%     if predict_label(i,1)>0
%         m=m+1;
%     end
%     if predict_label(i+50,1)<0
%         n=n+1;
%     end
% end
%
% s=m+n
% r=s/100
%  result 86%
% **********************************************************************
% 鏀寔鍚戦噺鏈猴紙Support Vector Machine, SVM锛?br>SVMStruct = svmtrain(train_data, train_label);
predict_label  = svmclassify(SVMStruct, test_data)
m=0;
n=0;
for i=1:50
    if predict_label(i,1)>0
        m=m+1;
    end
    if predict_label(i+50,1)<0
        n=n+1;
    end
end
 
s=m+n
r=s/100
 
%  result 86%

 

以上是关于matlab鑷甫鍚勭鍒嗙被鍣ㄧ殑浣跨敤绀轰緥的主要内容，如果未能解决你的问题，请参考以下文章

绐ユ帰婧愮爜锛岃鎴戞洿鍔犱紭闆呯殑浣跨敤Kafka鐢熶骇鑰咃紒

浠庢簮鐮佸垎鏋愬浣曚紭闆呯殑浣跨敤 Kafka 鐢熶骇鑰?/h1>

涓€瀛﹀氨浼氱殑git鍛戒护

姝ｅ垯琛ㄨ揪寮忕殑浣跨敤

ReactNative: 浣跨敤閫夋嫨鍣ㄧ粍浠禤icker缁勪欢

銆愯浆銆戙€怱alesfoece銆慉pproval Process 鍦?Apex 涓殑浣跨敤