OpenCV n fold_cross_validation - eiichiromomma/CVMLAB GitHub Wiki
(OpenCV) n-fold Cross Validation
OpenCV 機械学習で交差確認(n-fold Cross Validation)を行なう。C++の継承の復習ついでに作ってみたら便利だったので公開。
CrossVal class
概要
CrossValが基底となり、継承によりRandom TreesとSupport Vector MachineとNeural Networkについて交差確認が行なえるように作った。
他の機械学習についても仮想関数をオーバーライドすれば実装可能(な筈)。
大した事は行なっていないのでソースを読めば分かると思う。
前提
- Classificationのみ対象として作成
- Regressionについては各クラスのprintとCrossVal::classifier()の結果の評価方法に手を加える必要がある
- responsesは0から始まる
- 途中で面倒になってきたので実装は汚ない
- スペルミスや文法間違いは御愛嬌
使い方
共通事項
- 特徴量、responsesについては通常の機械学習と同様に作成
- CvTermCriteriaおよび各分類器のクラスを作成しておく
- 最後の引数はfold数ではなく、1回のfoldに使うサンプル数
- 結果はtestについての正誤テーブルと全体の正答率を表示
流れ
- CvTermCriteria作成
- 分類器作成
- 該当するCrossValの派生クラス作成
- classfier()呼び出し
Random Treesの場合
コンストラクタは
CrossValRTrees(CvRTreesへのポインタ, データへのポインタ, responsesへのポインタ, CvTermCriteria, CvRTParams, foldに使うサンプル数)
となり下のような使い方となる。
CvTermCriteria ctc = cvTermCriteria(CV_TERMCRIT_ITER,30,0.01);
CvRTrees* rt = new CvRTrees();
CrossValRTrees* cvrt = new CrossValRTrees(rt, data, responses, ctc,CvRTParams(15, 2, 0, false, data->cols, 0, false, 8, ctc.max_iter, ctc.epsilon, ctc.type),2000);
cvrt->classifier();
この場合、20,000件のdataに対して2,000個ごとのfoldとなり10-foldとなる。(以下同様)
MLP (Neural Network)の場合
コンストラクタは
CrossValMLP(CvANN_MLPへのポインタ, データへのポインタ, responsesへのポインタ, layer_sizes, CvANN_MLP_TrainParams, ニューロンの関数, foldに使うサンプル数)
となり下のような使い方となる。
CvTermCriteria ctc = cvTermCriteria(CV_TERMCRIT_ITER,300,0.1);
CvANN_MLP* m = new CvANN_MLP();
int layer_sz[] = { data->cols, 50, class_count };
CvMat layer_sizes = cvMat( 1, (int)(sizeof(layer_sz)/sizeof(layer_sz[0])), CV_32S, layer_sz );
CvANN_MLP_TrainParams tr_params = CvANN_MLP_TrainParams(ctc,CvANN_MLP_TrainParams::RPROP,0.3);
CrossValMLP* cvmlp = new CrossValMLP(m,data,responses,&layer_sizes,tr_params,CvANN_MLP::SIGMOID_SYM,4000);
cvmlp->classifier();
Support Vector Machineの場合
コンストラクタは
CrossValSvm(CvSVMへのポインタ, データへのポインタ, responsesへのポインタ, CvSVMParams, foldに使うサンプル数)
となり下のような使い方となる。
CvTermCriteria ctc = cvTermCriteria(CV_TERMCRIT_ITER,100,0.001);
CvSVM* s = new CvSVM();
CrossValSvm* cvsvm = new CrossValSvm(s,data, responses,
CvSVMParams(CvSVM::NU_SVC, CvSVM::RBF, 0.0, 0.1, 0.0, 0.0, 0.2, 0.0,NULL,ctc),4000);
cvsvm->classifier();
サンプル
sampleで配布されているletter_recog.cppへ適用。 letter_recog.cpp、myCrossVal.cpp、myCrossVal.h、letter-recognition.dataを同一フォルダに置き、Windowsの場合はdata以外をプロジェクトに加えてビルド。
Linux等の場合はMakefileでそれぞれコンパイルさせてリンクさせる。(動作未確認)
試行錯誤する方法としてSTLのlistを使ってみた。これも実装は汚ない。
配布ファイル
以下各結果
Random Treesの場合 その1
Training classifier: 1
Cross Validation: classifier 1
test1/10: Trees :23
Recognition Rate: train= 97.4%, test= 92.8%
Training classifier: 2
Cross Validation: classifier 2
test2/10: Trees :24
Recognition Rate: train= 97.4%, test= 92.3%
Training classifier: 3
Cross Validation: classifier 3
test3/10: Trees :19
Recognition Rate: train= 97.2%, test= 92.5%
Training classifier: 4
Cross Validation: classifier 4
test4/10: Trees :24
Recognition Rate: train= 97.2%, test= 93.4%
Training classifier: 5
Cross Validation: classifier 5
test5/10: Trees :22
Recognition Rate: train= 97.5%, test= 92.6%
Training classifier: 6
Cross Validation: classifier 6
test6/10: Trees :23
Recognition Rate: train= 97.5%, test= 93.2%
Training classifier: 7
Cross Validation: classifier 7
test7/10: Trees :20
Recognition Rate: train= 97.5%, test= 92.3%
Training classifier: 8
Cross Validation: classifier 8
test8/10: Trees :24
Recognition Rate: train= 97.3%, test= 92.0%
Training classifier: 9
Cross Validation: classifier 9
test9/10: Trees :25
Recognition Rate: train= 96.8%, test= 92.5%
Training classifier: 10
Cross Validation: classifier 10
test10/10: Trees :22
Recognition Rate: train= 97.4%, test= 92.0%
=======================Random Trees==========================
2000/20000CV,max_iter=30,epsilon=0.010000,max_depth=15,min_sample_count=2,max_categories=16,nactive_
vars=8,max_tree_count=30,forest_accuracy=0.010000,termcrit_type=1,
Number of Contents:789,766,736,805,768,775,773,734,755,747,739,761,792,783,753,803,783,758,748,796,813,764,752,787,786,734,
767 1 2 2 1 0 1 1 0 0 1 1 0 1 1 0 0 1 3 0 0 0 0 1 5 0
0 726 0 3 3 1 4 3 0 0 2 0 0 0 0 0 1 14 3 1 1 0 0 3 1 0
0 0 683 0 16 1 16 1 0 0 2 2 0 0 7 0 4 1 0 0 1 0 1 1 0 0
0 9 0 759 1 0 1 16 0 1 1 0 0 4 4 0 3 4 1 0 0 0 0 1 0 0
0 3 4 0 722 3 13 0 0 0 2 2 0 0 0 0 4 0 5 0 0 0 0 6 0 4
0 18 1 4 3 718 2 2 1 0 1 0 1 1 0 9 0 0 2 6 0 0 1 1 4 0
1 14 4 10 9 5 696 4 0 2 0 0 2 0 3 1 7 3 5 0 2 0 1 4 0 0
1 37 1 25 1 2 9 582 0 0 19 0 0 0 10 4 0 30 2 1 4 1 0 1 3 1
0 13 0 4 2 7 1 0 689 10 0 1 0 0 1 5 1 1 5 2 0 0 0 9 0 4
2 8 0 7 3 2 0 4 13 663 3 2 0 2 4 1 3 0 11 1 0 0 0 14 1 3
1 3 0 5 4 0 4 25 0 0 657 0 1 0 0 0 0 23 2 0 3 0 0 11 0 0
0 6 1 0 11 1 12 3 0 0 3 703 0 0 0 0 5 7 4 0 0 0 0 5 0 0
4 6 0 0 2 0 6 1 0 0 2 0 763 4 0 0 1 0 0 0 0 0 3 0 0 0
1 8 0 10 0 0 1 9 0 0 0 1 6 722 11 0 0 7 0 0 3 3 1 0 0 0
0 16 2 19 0 0 4 2 0 1 0 0 0 0 678 3 13 6 1 0 4 1 0 3 0 0
0 10 0 4 3 30 7 2 1 0 0 1 0 1 2 736 1 1 0 1 0 0 0 1 2 0
0 11 0 3 1 0 4 1 1 0 0 1 0 0 22 1 727 4 2 0 0 0 2 0 2 1
0 27 0 7 0 1 1 0 0 0 9 1 1 3 1 0 0 706 0 0 0 0 0 1 0 0
1 26 0 2 7 7 4 4 0 1 0 1 0 0 0 0 5 4 679 0 0 0 0 3 0 4
1 3 2 3 3 6 7 1 0 0 5 1 0 0 0 0 0 5 7 743 0 1 0 2 5 1
2 6 0 2 0 0 3 7 0 1 0 0 6 9 8 0 0 0 1 0 766 2 0 0 0 0
0 29 0 0 1 1 2 0 0 0 0 0 1 1 3 1 1 3 0 0 1 711 5 0 4 0
0 1 0 0 0 0 1 3 0 0 0 0 7 1 3 0 1 1 0 0 3 0 731 0 0 0
0 12 0 3 7 3 0 6 2 0 9 1 0 0 1 0 0 3 1 0 0 0 0 739 0 0
0 1 0 4 0 0 0 0 0 0 0 0 3 1 1 1 1 0 1 6 2 6 0 1 758 0
0 3 0 8 9 1 1 0 1 1 0 1 0 0 0 0 8 2 8 1 0 0 0 1 0 689
Total Recognition rate: train = 97.3%, test = 92.6%
Random Treesの場合 その2
Training classifier: 1
Cross Validation: classifier 1
test1/10: Trees :17
Recognition Rate: train= 97.5%, test= 92.4%
Training classifier: 2
Cross Validation: classifier 2
test2/10: Trees :23
Recognition Rate: train= 97.2%, test= 91.3%
Training classifier: 3
Cross Validation: classifier 3
test3/10: Trees :21
Recognition Rate: train= 97.1%, test= 92.0%
Training classifier: 4
Cross Validation: classifier 4
test4/10: Trees :21
Recognition Rate: train= 97.6%, test= 92.3%
Training classifier: 5
Cross Validation: classifier 5
test5/10: Trees :26
Recognition Rate: train= 97.1%, test= 93.0%
Training classifier: 6
Cross Validation: classifier 6
test6/10: Trees :24
Recognition Rate: train= 97.5%, test= 91.8%
Training classifier: 7
Cross Validation: classifier 7
test7/10: Trees :26
Recognition Rate: train= 97.0%, test= 92.4%
Training classifier: 8
Cross Validation: classifier 8
test8/10: Trees :21
Recognition Rate: train= 97.1%, test= 91.7%
Training classifier: 9
Cross Validation: classifier 9
test9/10: Trees :27
Recognition Rate: train= 97.3%, test= 92.1%
Training classifier: 10
Cross Validation: classifier 10
test10/10: Trees :19
Recognition Rate: train= 97.3%, test= 93.6%
=======================Random Trees==========================
2000/20000CV,max_iter=30,epsilon=0.010000,max_depth=15,min_sample_count=3,max_categories=16,nactive_
vars=8,max_tree_count=30,forest_accuracy=0.010000,termcrit_type=1,
Number of Contents:789,766,736,805,768,775,773,734,755,747,739,761,792,783,753,803,783,758,748,796,813,764,752,787,786,734,
766 3 2 2 1 0 1 0 0 0 3 2 0 0 0 0 0 1 1 0 0 0 0 3 4 0
0 723 0 1 3 1 4 3 0 0 0 0 0 0 2 2 0 19 4 0 0 2 0 2 0 0
0 0 685 0 12 2 16 0 0 0 1 3 0 0 9 1 2 1 1 0 3 0 0 0 0 0
0 15 0 745 1 0 1 12 2 0 1 0 0 4 7 2 1 12 0 0 0 0 0 1 0 1
0 1 7 0 705 2 18 0 0 0 3 4 0 0 0 0 4 3 6 1 1 0 0 8 0 5
0 21 0 6 5 708 0 2 1 1 0 0 0 1 0 16 0 1 4 6 0 0 1 1 1 0
1 13 2 11 4 0 709 4 0 2 2 0 1 0 4 1 4 7 0 0 0 1 1 6 0 0
1 26 1 28 3 0 8 582 0 0 23 0 1 0 12 4 2 36 2 0 2 0 0 2 1 0
0 17 1 6 2 11 0 0 688 6 0 1 0 0 0 2 5 2 2 1 0 0 0 9 0 2
0 14 0 2 4 7 0 6 10 674 1 0 0 0 1 0 3 4 8 0 1 0 0 10 1 1
0 2 1 5 2 1 5 21 0 0 652 0 2 1 0 0 0 32 1 0 2 0 0 12 0 0
0 7 2 0 9 1 11 1 0 0 2 702 0 0 2 2 2 7 3 0 0 0 0 9 0 1
1 9 0 0 1 1 2 1 0 0 2 0 758 5 2 1 0 3 0 0 0 0 5 0 1 0
1 5 0 14 0 1 0 10 0 0 0 0 7 717 6 0 0 15 0 0 1 3 1 2 0 0
0 17 2 22 1 0 1 2 0 0 0 0 1 1 669 2 11 13 2 0 4 0 0 5 0 0
0 10 0 2 8 30 3 3 0 1 1 0 0 0 2 733 0 2 0 1 0 1 3 1 2 0
1 9 0 7 2 0 7 0 1 0 0 1 0 0 21 2 723 4 1 0 1 1 0 0 0 2
0 31 0 6 1 0 1 3 0 0 9 0 1 3 0 1 0 698 1 0 0 1 0 2 0 0
1 22 0 1 11 5 4 5 0 2 0 2 0 0 1 1 2 4 672 1 0 0 0 6 1 7
0 3 6 2 1 7 3 1 0 0 8 1 0 0 1 0 1 1 4 752 0 1 0 0 4 0
0 2 2 2 0 0 1 8 0 0 2 0 10 6 12 0 2 0 0 0 765 0 1 0 0 0
1 24 1 1 0 1 2 2 0 0 0 0 1 0 3 4 0 3 0 0 2 708 5 0 6 0
0 3 0 0 0 0 1 1 0 0 0 0 4 0 7 1 1 1 0 0 4 1 727 0 1 0
0 7 0 5 9 2 0 3 0 1 7 1 0 0 1 0 0 0 1 0 0 0 0 750 0 0
0 4 2 2 0 0 1 1 0 0 0 0 0 0 0 1 3 0 2 8 3 6 4 1 748 0
1 1 0 4 8 0 0 0 0 6 0 1 0 0 0 0 9 3 7 1 0 0 0 1 0 692
Total Recognition rate: train = 97.3%, test = 92.3%
Support Vector Machineの場合 その1
Training classifier: 1
Cross Validation: classifier 1
test1/10: 9359 SupportVectors,Recognition Rate: train= 99.6%, test= 97.4%
Training classifier: 2
Cross Validation: classifier 2
test2/10: 9336 SupportVectors,Recognition Rate: train= 99.7%, test= 97.0%
Training classifier: 3
Cross Validation: classifier 3
test3/10: 9312 SupportVectors,Recognition Rate: train= 99.7%, test= 97.0%
Training classifier: 4
Cross Validation: classifier 4
test4/10: 9310 SupportVectors,Recognition Rate: train= 99.7%, test= 97.4%
Training classifier: 5
Cross Validation: classifier 5
test5/10: 9352 SupportVectors,Recognition Rate: train= 99.7%, test= 97.5%
Training classifier: 6
Cross Validation: classifier 6
test6/10: 9383 SupportVectors,Recognition Rate: train= 99.7%, test= 97.7%
Training classifier: 7
Cross Validation: classifier 7
test7/10: 9315 SupportVectors,Recognition Rate: train= 99.7%, test= 97.0%
Training classifier: 8
Cross Validation: classifier 8
test8/10: 9370 SupportVectors,Recognition Rate: train= 99.6%, test= 97.4%
Training classifier: 9
Cross Validation: classifier 9
test9/10: 9332 SupportVectors,Recognition Rate: train= 99.7%, test= 97.2%
Training classifier: 10
Cross Validation: classifier 10
test10/10: 9344 SupportVectors,Recognition Rate: train= 99.7%, test= 96.7%
=================Supoprt Vector Machine=======================
NU_SVC,RBF,2000/20000CV,max_iter=300,epsilon=0.001000,gamma=0.100000,nu=0.200000,
Number of Contents:789,766,736,805,768,775,773,734,755,747,739,761,792,783,753,803,783,758,748,796,813,764,752,787,786,734,
784 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0
0 745 0 3 2 0 0 3 0 0 0 0 1 0 0 0 0 5 1 0 1 4 0 1 0 0
0 0 712 0 7 0 3 1 0 0 0 1 2 0 7 0 0 0 0 1 0 1 1 0 0 0
1 1 0 782 0 0 0 9 0 0 0 0 2 5 5 0 0 0 0 0 0 0 0 0 0 0
0 3 1 0 748 1 4 0 0 0 0 3 0 0 0 1 2 0 0 0 0 0 0 0 0 5
0 1 0 1 2 747 0 1 0 2 0 0 0 0 0 13 0 0 0 6 0 1 0 1 0 0
0 1 3 10 6 0 747 0 0 0 0 0 1 0 1 0 0 0 1 0 0 1 2 0 0 0
0 3 1 19 1 0 6 679 0 1 10 0 1 0 1 0 0 10 0 0 2 0 0 0 0 0
0 0 1 0 0 3 0 0 719 29 0 0 0 1 0 1 0 0 0 0 0 0 0 1 0 0
0 0 0 2 0 1 0 2 21 716 0 0 0 2 1 0 1 0 0 0 1 0 0 0 0 0
0 0 0 2 0 0 0 10 0 0 705 0 0 0 0 0 0 13 0 0 1 0 0 8 0 0
0 1 0 0 4 0 2 3 0 1 0 745 1 0 0 0 0 2 0 0 0 0 0 2 0 0
0 5 0 0 0 0 2 0 0 0 0 0 782 0 0 0 0 0 0 0 1 1 1 0 0 0
0 1 0 4 0 0 0 3 0 0 0 0 5 758 6 0 0 5 0 0 0 1 0 0 0 0
0 1 4 8 0 0 0 0 0 0 0 0 4 0 729 0 3 0 0 0 1 0 3 0 0 0
0 2 0 0 1 17 0 1 0 0 0 1 0 0 0 774 4 0 0 0 0 0 1 0 2 0
1 0 0 1 0 0 0 0 0 0 0 0 3 0 5 3 769 1 0 0 0 0 0 0 0 0
0 14 0 0 0 0 0 5 0 0 8 0 0 5 0 0 2 724 0 0 0 0 0 0 0 0
0 1 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 1 743 0 0 0 0 0 0 0
0 1 2 4 0 0 0 1 0 0 0 0 0 0 0 1 0 1 0 779 0 1 0 1 5 0
1 0 0 0 0 0 0 3 0 0 0 0 3 0 0 0 0 0 0 0 805 1 0 0 0 0
0 17 0 0 0 2 1 0 0 0 0 0 2 1 0 2 0 0 0 0 0 737 1 0 1 0
0 0 0 0 0 0 1 0 0 0 0 0 5 0 1 0 0 0 0 0 1 0 744 0 0 0
0 0 0 3 3 0 0 0 0 0 4 0 1 0 2 0 1 1 0 1 0 0 0 770 1 0
0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 1 0 1 1 780 0
0 0 0 0 1 0 0 0 0 0 0 0 4 0 0 0 7 0 0 0 0 0 0 0 0 722
Total Recognition rate: train = 99.7%, test = 97.2%
Support Vector Machineの場合 その2
Training classifier: 1
Cross Validation: classifier 1
test1/10: 9320 SupportVectors,Recognition Rate: train= 97.3%, test= 74.6%
Training classifier: 2
Cross Validation: classifier 2
test2/10: 9280 SupportVectors,Recognition Rate: train= 98.7%, test= 75.8%
Training classifier: 3
Cross Validation: classifier 3
test3/10: 9332 SupportVectors,Recognition Rate: train= 97.8%, test= 73.0%
Training classifier: 4
Cross Validation: classifier 4
test4/10: 9253 SupportVectors,Recognition Rate: train= 98.3%, test= 77.3%
Training classifier: 5
Cross Validation: classifier 5
test5/10: 9299 SupportVectors,Recognition Rate: train= 96.3%, test= 73.0%
Training classifier: 6
Cross Validation: classifier 6
test6/10: 9327 SupportVectors,Recognition Rate: train= 98.9%, test= 77.4%
Training classifier: 7
Cross Validation: classifier 7
test7/10: 9356 SupportVectors,Recognition Rate: train= 99.7%, test= 78.5%
Training classifier: 8
Cross Validation: classifier 8
test8/10: 9308 SupportVectors,Recognition Rate: train= 99.4%, test= 77.6%
Training classifier: 9
Cross Validation: classifier 9
test9/10: 9352 SupportVectors,Recognition Rate: train= 98.8%, test= 78.0%
Training classifier: 10
Cross Validation: classifier 10
test10/10: 9303 SupportVectors,Recognition Rate: train= 98.4%, test= 75.4%
=================Supoprt Vector Machine=======================
C_SVC,RBF,2000/20000CV,max_iter=300,epsilon=0.001000,gamma=3.000000,C=1000.000000,
Number of Contents:789,766,736,805,768,775,773,734,755,747,739,761,792,783,753,803,783,758,748,796,813,764,752,787,786,734,
661 0 0 0 0 0 0 0 0 0 43 0 27 0 0 58 0 0 0 0 0 0 0 0 0 0
0 599 0 1 1 0 0 1 0 1 34 0 51 0 0 57 0 9 1 0 0 10 0 1 0 0
0 0 596 0 1 0 1 0 0 0 40 0 42 0 0 55 0 0 0 1 0 0 0 0 0 0
0 1 0 646 0 0 0 8 0 0 41 0 38 2 2 64 0 2 0 0 0 0 0 0 1 0
0 1 4 0 589 0 6 1 0 0 61 3 42 0 0 59 0 0 0 0 0 0 0 1 0 1
0 0 0 0 0 591 0 0 0 1 41 0 42 0 0 95 0 0 1 3 0 0 0 0 1 0
0 1 3 2 4 0 581 0 0 0 56 0 55 0 0 68 2 0 0 0 0 0 1 0 0 0
0 6 0 13 2 0 3 468 0 1 82 0 56 5 2 90 0 5 0 0 1 0 0 0 0 0
0 0 0 0 0 1 0 0 569 28 42 0 55 0 0 60 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 1 2 596 34 0 45 1 0 68 0 0 0 0 0 0 0 0 0 0
0 1 0 0 4 0 0 13 0 0 626 0 38 0 0 46 0 5 0 0 0 0 0 6 0 0
0 0 0 0 3 0 0 0 0 2 53 593 53 0 0 57 0 0 0 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0 0 0 40 0 701 0 0 48 0 0 0 0 0 2 0 0 0 0
0 0 0 4 0 0 0 1 0 0 38 0 66 610 1 58 0 4 0 0 0 1 0 0 0 0
0 0 2 9 0 0 0 2 0 0 55 0 71 6 524 72 10 0 0 0 2 0 0 0 0 0
0 0 0 0 0 22 0 0 0 0 131 0 133 0 0 515 2 0 0 0 0 0 0 0 0 0
0 0 1 0 0 0 2 0 0 0 85 0 33 0 8 121 532 0 0 0 0 0 0 0 0 1
0 5 0 2 0 0 0 5 0 0 52 1 45 4 0 81 0 562 1 0 0 0 0 0 0 0
0 1 0 0 1 0 0 0 0 0 39 0 47 0 0 62 0 0 598 0 0 0 0 0 0 0
0 0 3 0 0 7 0 0 0 0 54 0 27 0 0 52 0 0 0 643 0 1 0 1 8 0
0 0 0 0 0 0 0 2 0 0 32 0 44 1 0 58 0 0 0 0 675 0 0 0 1 0
0 5 0 0 0 0 0 0 0 0 51 0 58 1 0 80 0 0 0 0 0 566 1 0 2 0
0 0 0 0 0 0 0 0 0 0 48 0 54 0 0 65 0 0 0 0 0 0 585 0 0 0
0 0 0 0 5 0 0 1 0 0 96 0 44 0 0 94 0 0 0 0 0 0 0 547 0 0
0 0 0 0 0 0 0 0 0 0 37 0 99 0 0 127 0 0 0 17 1 6 0 0 499 0
0 0 0 0 4 0 0 0 0 0 58 0 54 0 0 73 0 0 1 0 0 0 0 0 0 544
Neural Networkの場合 その1
Training classifier: 1
Cross Validation: classifier 1
test1/5: 3 Layers,Recognition Rate: train= 85.2%, test= 83.0%
Training classifier: 2
Cross Validation: classifier 2
test2/5: 3 Layers,Recognition Rate: train= 83.1%, test= 81.2%
Training classifier: 3
Cross Validation: classifier 3
test3/5: 3 Layers,Recognition Rate: train= 85.9%, test= 84.2%
Training classifier: 4
Cross Validation: classifier 4
test4/5: 3 Layers,Recognition Rate: train= 85.8%, test= 84.9%
Training classifier: 5
Cross Validation: classifier 5
test5/5: 3 Layers,Recognition Rate: train= 86.9%, test= 86.1%
Layer = {16, 50, 26, }
Number of Contents:789,766,736,805,768,775,773,734,755,747,739,761,792,783,753,803,783,758,748,796,813,764,752,787,786,734,
721 1 0 1 0 4 3 2 0 6 8 4 6 4 1 1 0 3 2 0 12 3 1 3 2 1
2 692 0 8 9 2 4 8 0 1 4 0 0 0 0 0 0 18 8 1 3 0 0 6 0 0
1 3 591 1 34 1 19 10 0 0 28 2 0 1 14 0 5 3 2 0 15 0 0 4 2 0
0 14 0 726 1 2 3 7 1 0 5 2 2 6 11 1 0 7 4 1 7 0 0 5 0 0
0 5 5 1 668 5 23 0 1 0 10 1 0 0 0 0 7 0 19 1 0 0 0 14 0 8
0 14 0 6 11 659 1 0 7 7 0 0 0 1 1 20 0 5 19 11 0 2 2 6 2 1
2 22 18 5 28 7 544 2 0 0 16 1 3 2 34 8 34 15 11 0 9 0 0 10 1 1
26 15 0 35 13 23 11 290 0 7 73 7 3 36 52 15 8 52 8 1 26 6 1 21 5 0
0 7 0 13 1 15 1 0 665 8 1 0 0 0 2 6 1 2 22 0 0 0 0 9 0 2
3 4 0 2 4 30 0 3 12 635 0 0 0 2 7 0 2 3 15 0 3 0 0 18 1 3
4 5 4 4 12 1 2 8 0 0 631 0 2 3 3 0 0 23 5 0 1 0 0 31 0 0
1 10 0 0 20 2 18 4 0 0 3 656 0 1 0 0 3 9 10 0 5 0 0 16 2 1
68 13 0 4 1 0 2 3 0 0 2 0 668 9 5 0 0 5 2 0 3 0 7 0 0 0
27 5 0 19 0 13 0 37 0 0 1 0 5 606 23 1 0 14 1 3 13 6 8 1 0 0
4 0 4 14 0 0 9 3 0 3 1 0 2 3 650 4 14 15 4 0 8 0 13 2 0 0
1 10 0 5 1 26 13 1 2 1 2 0 0 0 7 711 8 1 3 0 0 1 1 0 9 0
11 12 0 7 14 0 16 0 5 0 7 5 0 0 20 0 646 8 14 0 1 0 2 1 0 14
7 36 0 13 4 7 7 5 0 0 21 2 3 10 1 1 3 625 6 0 5 0 0 2 0 0
3 25 5 6 26 20 5 0 15 7 3 3 0 1 3 1 7 11 542 6 2 1 0 23 5 28
0 14 0 6 19 3 8 10 0 0 5 1 0 0 2 2 4 5 8 673 8 4 0 9 7 8
2 8 1 2 0 0 1 6 0 2 3 0 12 4 9 0 0 2 0 0 759 1 1 0 0 0
0 20 0 3 3 4 11 1 0 0 2 0 0 1 8 2 0 3 2 1 2 680 11 0 10 0
8 15 0 1 2 1 3 2 0 0 1 0 5 5 17 0 4 6 0 1 8 3 669 1 0 0
0 6 0 8 14 2 0 0 0 1 13 0 0 0 5 1 0 2 3 2 1 0 0 722 6 1
2 4 0 2 0 4 3 5 0 1 1 0 0 1 3 11 11 0 4 8 7 11 2 3 698 5
4 0 0 5 12 4 1 0 2 3 0 10 0 0 1 0 3 4 27 2 1 0 0 7 1 647
Total Recognition rate: train = 85.4%, test = 83.9%
Neural Networkの場合 その2
Training classifier: 1
Cross Validation: classifier 1
test1/5: 4 Layers,Recognition Rate: train= 85.8%, test= 84.6%
Training classifier: 2
Cross Validation: classifier 2
test2/5: 4 Layers,Recognition Rate: train= 88.1%, test= 86.3%
Training classifier: 3
Cross Validation: classifier 3
test3/5: 4 Layers,Recognition Rate: train= 86.5%, test= 84.6%
Training classifier: 4
Cross Validation: classifier 4
test4/5: 4 Layers,Recognition Rate: train= 87.4%, test= 84.9%
Training classifier: 5
Cross Validation: classifier 5
test5/5: 4 Layers,Recognition Rate: train= 86.8%, test= 84.2%
Layer = {16, 25, 25, 26, }
Number of Contents:789,766,736,805,768,775,773,734,755,747,739,761,792,783,753,803,783,758,748,796,813,764,752,787,786,734,
704 3 0 5 0 4 0 2 0 3 7 2 2 3 3 1 3 10 8 0 9 3 3 5 5 4
0 658 0 15 2 1 3 2 0 0 5 3 0 0 6 1 3 44 12 0 0 1 0 8 1 1
3 0 592 0 26 4 25 13 0 2 21 3 3 3 11 0 9 2 2 1 11 1 1 3 0 0
2 9 1 723 1 3 0 13 0 0 1 0 3 6 8 2 1 16 7 0 6 0 0 1 1 1
0 7 9 0 646 10 25 2 1 1 9 9 1 0 0 0 11 6 7 0 1 0 0 13 0 10
0 13 5 4 20 633 3 7 2 3 0 1 1 3 0 24 0 4 22 12 0 4 2 5 6 1
3 7 16 7 8 4 579 7 2 3 15 6 1 1 20 6 20 17 16 0 5 12 1 11 1 5
3 17 5 29 1 8 7 523 0 1 29 5 2 15 16 4 8 34 0 1 6 7 0 6 6 1
1 3 3 10 6 14 0 0 657 6 0 3 0 0 2 6 0 0 17 1 0 0 0 16 4 6
6 1 0 10 4 18 0 4 10 599 2 3 0 2 3 6 2 6 20 2 4 0 0 23 3 19
2 5 1 7 9 2 5 19 0 2 603 1 2 8 1 0 0 37 1 2 1 4 2 24 1 0
0 6 4 1 24 1 15 5 1 1 8 650 0 0 0 0 2 7 10 1 7 0 0 15 2 1
1 14 0 0 0 0 8 3 0 0 0 3 727 20 2 0 0 7 0 0 4 0 3 0 0 0
5 6 1 12 0 1 1 16 0 1 2 0 7 685 16 5 0 14 0 0 1 2 6 1 1 0
2 3 3 23 0 0 6 7 0 2 4 1 2 12 621 3 21 16 5 0 2 6 13 1 0 0
1 9 0 2 3 35 7 1 2 0 4 1 1 0 3 708 6 6 3 1 0 1 1 0 8 0
3 15 1 4 8 0 10 0 2 0 2 4 1 1 31 0 652 3 16 0 3 5 2 3 4 13
5 37 0 12 2 0 6 21 0 0 15 2 1 6 3 0 2 634 0 0 1 1 0 8 2 0
3 39 1 8 14 13 4 0 4 4 0 2 0 0 6 0 5 5 599 3 2 1 0 9 5 21
0 7 1 4 14 7 6 12 0 0 7 9 0 0 3 2 8 5 17 653 3 4 0 16 8 10
1 4 4 0 0 1 3 15 0 0 2 2 11 5 7 1 6 4 0 0 736 2 6 3 0 0
0 20 0 0 0 0 4 4 0 1 0 0 2 3 2 5 2 7 1 0 0 690 12 1 10 0
2 8 0 0 0 0 4 5 0 1 0 0 8 17 8 0 0 8 0 0 7 7 677 0 0 0
0 6 1 8 16 3 2 5 0 1 11 1 0 0 1 0 2 3 18 4 1 0 0 692 7 5
1 4 1 5 0 8 0 3 0 0 0 6 2 2 1 12 9 2 5 12 6 13 1 4 687 2
4 3 0 6 16 1 0 1 0 1 0 2 0 0 0 0 5 2 33 1 0 0 0 7 2 650
Total Recognition rate: train = 86.9%, test = 84.9%