KESIMPULAN DAN SARAN
DAFTAR PUSTAKA
García-Pedrajas, N. & Ortiz-Boyer, D. September 2009. Boosting K-Nearest Neighbor Classifier By Means Of Input Space Projection. Expert System With Application 37(7): pp.10570-10582.
Gou, J., Yi. Z., Du. L. & Xiong, T. January 2012. A Local Mean-Based k-Nearest Centroid Neighbor Classifier. The Computer Journal 55(6) : pp. 1058-1071.
Han, J., Kamber, M. & Pei, J. 2011. Data Mining: Concepts and Techniques. 3rd Edition. Morgan Kaufmann : Amsterdam.
Iswarya, P. & Radha, V. March 2015. Ensemble learning approach in Improved K Nearest Neighbor algorithm for Text Categorization. 2015 International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS), pp. 1-5.
Jabbar, M.A., Deekshatulu, B.L. & Chandra. P. January 2013. Classification of Heart Disease Using K- Nearest Neighbor and Genetic Algorithm. International Conference on Computational Intelligence: Modeling Techniques and Applications (CIMTA), pp. 85-94.
Jo, T. March 2017. Using K Nearest Neighbors for Text Segmentation with Feature Similarity. International Conference on Communication, Control, Computing and Electronics Engineering (ICCCCEE), pp. 1-5.
Kalaivani, P. & Shunmuganathan, K.L. March 2014. An Improved K-Nearest-Neighbor Algorithm Usinggenetic Algorithm For Sentiment Classification. 2014 International Conference on Circuit, Power and Computing Technologies (ICCPCT), pp. 1647-1651.
Kataria, A. & Singh, M.D. June 2013. A Review Data Classification Using K-Nearest Neighbour Algorithm. International Journal of Emerging Technology and Advanced Engineering 3(6): 354-360.
Kuhkan, M. June 2016. A Method to Improve the Accuracy of K - Nearest Neighbor Algorithm. Internatonal Journal of Computer Engineering and Information Technology 8(6): 90-95.
Lei, Z., Wang, S. & Xu, D. February 2016. Protein Sub-cellular Localization Based on Noise-Intensity-Weighted Linear Discriminant Analysis and an Improved K-Nearest-Neighbor Classifier. 2016 9th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI 2016), pp. 1871-1876.
Lidya, S.K., Sitompul, O.S. & Efendi, S. March 2015. Sentiment Analysis Pada Teks Bahasa Indonesia Menggunakan Support Vector Machine (SVM) Dan K-Nearest Neighbor (K-NN). Seminar Nasional Teknologi Informasi dan Komunikasi 2015, pp. 1-8.
Loohach, R. & Garg, K. July 2012. Effect of Distance Functions on K-Means Clustering Algorithm. International Journal of Computer Applications 49(6): 7-9.Prasetyo, E. (2014). Data Mining-Mengolah Data Menjadi Informasi Menggunakan MATLAB. Yogyakarta: ANDI.
Mitani, Y. & Hamamoto, Y. July 2006. A Local Mean-Based Nonparametric Classifier. Patern Recognition Letter 27(10): 1151-1159.
Nababan, A A., Sitompul, O.S., & Tulus. April 2018. Attribute Weighting Based K-Nearest Neighbor Using Gain Ratio. 2018. IOP Conf. Series: Journal of Physics: Conf. Series 1007 012007.
Ougiaroglou, S. & Evangelidis, G. October 2012. Fast and Accuratek-Nearest Neighbor Classification using Prototype Selection by Clustering. Panhellenic Conference on Informatics, pp. 168-173.
Pan, Z., Wang, Y. & Ku, W. September 2016. A New K-Harmonic Nearest Neighbor Classifer Based On The Multi-Local Means. Expert Systems With Applications 67: 115-125.
Pan, Z., Wang, Y. & Ku, W. January 2017. A New General Nearest Neighbor Classification Based On The Mutual Neighborhood Information. Knowledge-Based Systems 121: 142-152.
Pandit, S. & Gupta, S. A. December 2011. Comparative Study on Distance Measuring Approaches for Clustering. International Journal of Research in Computer Science 2(1): pp. 29-31.
Prasetyo, E. (2014). Data Mining-Mengolah Data Menjadi Informasi Menggunakan MATLAB. Yogyakarta: ANDI.
Rui-Jia, W. & Xing, W., 2014. Radar Emitter Recognition in Airborne RWR/ESM Based on Improved K Nearest Neighbor Algorithm. 2014 IEEE International Conference on Computer and Information Technology (CIT), pp. 148-151.
Sánchez, A.S., Iglesias-Rodríguez, F.J., Fernándes, P.R. & Juez, F.J.de.C. 2015.
Applying The K-Nearest Neighbor Technique To The Classification Of Workers According To Their Risk Of Suffering Musculoskeletal Disorders. International Journal of Indsutrial Ergonomics 52: 92-99.
Shirkhorshidi, A.S., Aghabozorgi, S. & Wah, T.Y. December 2015. A Comparison Study on Similarity and Dissimilarity Measures in Clustering Continuous Data.
PLUS ONE, pp. 1-20.
Syaliman. K.U., Sitompul. O.S. & Nababan, E.B. March 2017. Improving The Accuracy Of K-Nearest Neighbor Using Local Mean Based And Distance Weight. 2nd International Conference on Computing and Applied Informatics 2017. Pp. 1-5.
Song, Y., Liang, J., Lu, J. & Zhao, X. August 2017. An Efficient Instance Selection Algorithm For K Nearest Neighbor Regression. Neurocomputing 251: 26-34.
Wang. J., Neskovic . P. & Cooper L.N., January 2007. Improving Nearest Neighbor Rule With A Simple Adaptive Distance Measure. Pattern Recognition Letter 28:
207-213.
Zheng, K. Si, G. Diao, L. Zhou, Z. Chen, J. & Yue W., May 2017. Applications Of Support Vector Machine And Improved k-Nearest Neighbor Algorithm In Fault Diagnosis And fault Degree Evaluation Of Gas Insulated Switchgear. 1st International Conference on Electrical Materials and Power Equipment - Xi'an – China, pp. 364-368.
Lampiran Data Set. Link Dataset
Lampiran 1. Rincian Data Ionosphare
https://archive.ics.uci.edu/ml/datasets/ionosphare Lampiran 2. Rincian Data Wine
https://archive.ics.uci.edu/ml/datasets/wine Lampiran 3. Rincian Data Glass Identification
https://archive.ics.uci.edu/ml/datasets/glass+identification Lampiran 4. Rincian Data Haberman
https://archive.ics.uci.edu/ml/datasets/Haberman%27s+Survival Lampiran 5. Rincian Data Hayes-roth
https://archive.ics.uci.edu/ml/datasets/Hayes-Roth
Lampiran 6. Listing Program
Listing Program Dataset Ionosphare
while(metode1<metode3 || metode1<metode2) load ionosphere
load indexionfix.mat z=[1,2,3,4,5,6,7,8,9,10];
cp=knnclassify(test, train, trainc, z(x), 'euclidean');
AB{x,1} = classperf(testc, cp); %akurasi
if(ib~=0 && ig~=0)
acc(x)=A{x,1}.CorrectRate;
ace(x)=C{x,1}.CorrectRate;
end
metode1=sum(ace)/10;
metode2=sum(acc)/10;
metode3=sum(acd)/10;
end
disp(['Perbandingan Nilai Akurasi :']);
disp(['Nilai K K-NN Konvensioanl LMKNN Metode Yang
Listing Program Dataset Wine
disttmp(j,1)=sqrt(sum((test(i,:)-train(j,:)).^2));
end
dist(i,1)={disttmp};
end
for x=1:size(z,2)
%test K-NN konvensional
cp=knnclassify(test, train, trainc, z(x), 'euclidean');
AB{x,1} = classperf(testc, cp); %akurasi acd(x)=AB{x,1}.CorrectRate;
%---sort---for n=1:size(dist,1)
[B,index]=sort(dist{n,1});
%index=index(1:z(x),1);
distlma=sqrt(sum((test(n,:)-lma).^2));
distlmb=sqrt(sum((test(n,:)-lmb).^2));
distlmc=sqrt(sum((test(n,:)-lmc).^2));
distlha=1/wa;
distlhb=1/wb;
distlhc=1/wc;
if(ia~=0 && ib~=0 && ic~=0)
if(distlma<distlmb && distlma<distlmc) class(n,1)=1;
elseif (distlmb<distlma && distlmb<distlmc) class(n,1)=2;
else
class(n,1)=3;
end
if(distlha<distlhb && distlha<distlhc) class1(n,1)=1;
elseif(distlhb<distlha && distlhb<distlhc) class1(n,1)=2;
Listing Program Dataset Glass
disttmp(j,1)=sqrt(sum((test(i,:)-train(j,:)).^2));
end
dist(i,1)={disttmp};
end
for x=1:size(z,2)
%test K-NN konvensional
cp=knnclassify(test, train, trainc, z(x), 'euclidean');
AB{x,1} = classperf(testc, cp); %akurasi acd(x)=AB{x,1}.CorrectRate;
%---sort---for n=1:size(dist,1)
[B,index]=sort(dist{n,1});
%index=index(1:z(x),1);
elseif(trainc(index(i,1))==3 && ic~=z(x))
distlma=sqrt(sum((test(n,:)-lma).^2));
distlmb=sqrt(sum((test(n,:)-lmb).^2));
distlmc=sqrt(sum((test(n,:)-lmc).^2));
distlmd=sqrt(sum((test(n,:)-lmd).^2));
distlme=sqrt(sum((test(n,:)-lme).^2));
distlmg=sqrt(sum((test(n,:)-lmg).^2));
distlha=1/wa;
if(ia~=0 && ib~=0 && ic~=0 && id~=0 && ie~=0 && ig~=0)
if(distlma<distlmb && distlma<distlmc && distlma<distlmd &&
distlma<distlme && distlma<distlmg) class(n,1)=1;
elseif (distlmb<distlma && distlmb<distlmc && distlmb<distlmd
&& distlmb<distlme && distlmb<distlmg) class(n,1)=2;
elseif (distlmc<distlma && distlmc<distlmb && distlmc<distlmd
&& distlmc<distlme && distlmc<distlmg) class(n,1)=3;
elseif (distlmd<distlma && distlmd<distlmb && distlmd<distlmc
&& distlmd<distlme && distlmd<distlmg) class(n,1)=5;
elseif (distlme<distlma && distlme<distlmb && distlme<distlmc
&& distlme<distlmd && distlme<distlmg) class(n,1)=6;
else
class(n,1)=7;
end
if(distlha<distlhb && distlha<distlhc && distlha<distlhd &&
distlha<distlhe && distlha<distlhg) class1(n,1)=1;
elseif(distlhb<distlha && distlhb<distlhc && distlhb<distlhd
&& distlhb<distlhe && distlhb<distlhg)
class1(n,1)=2;
elseif(distlhc<distlha && distlhc<distlhb && distlhc<distlhd
&& distlhc<distlhe && distlhc<distlhg) class1(n,1)=3;
elseif(distlhd<distlha && distlhd<distlhb && distlhd<distlhc
&& distlhd<distlhe && distlhd<distlhg) class1(n,1)=5;
elseif(distlhe<distlha && distlhe<distlhb && distlhe<distlhc
&& distlhe<distlhd && distlhe<distlhg)
if(ib<ia && ic<ia && id<ia && ie<ia && ig<ia) class(n,1)=1;
class1(n,1)=1;
elseif(ia<ib && ic<ib && id<ib && ie<ib && ig<ib) class(n,1)=2;
class1(n,1)=2;
elseif(ia<ic && ib<ic && id<ic && ie<ic && ig<ic) class(n,1)=3;
class1(n,1)=3;
elseif(ia<id && ib<id && ic<id && ie<id && ig<id) class(n,1)=5;
class1(n,1)=5;
elseif(ia<ie && ib<ie && ic<ie && id<ie && ig<ie) class(n,1)=6;
Listing Program Dataset Haberman
disttmp(j,1)=sqrt(sum((test(i,:)-train(j,:)).^2));
end
dist(i,1)={disttmp};
end
for x=1:size(z,2)
%test K-NN konvensional
cp=knnclassify(test, train, trainc, z(x), 'euclidean');
AB{x,1} = classperf(testc, cp); %akurasi acd(x)=AB{x,1}.CorrectRate;
%---sort---for n=1:size(dist,1)
[B,index]=sort(dist{n,1});
%index=index(1:z(x),1);
elseif(strcmp(trainc(index(i,1)) ,'2') && ig~=z(x)) ig=ig+1;
distlmb=sqrt(sum((test(n,:)-lmb).^2));
distlmg=sqrt(sum((test(n,:)-lmg).^2));
distlhb=1/wb;
distlhg=1/wg;
if(ib~=0 && ig~=0) if(distlmb<distlmg) class(n,1)={'1'};
else
Listing Program Dataset Hayes-Roth
disttmp(j,1)=sqrt(sum((test(i,:)-train(j,:)).^2));
end
dist(i,1)={disttmp};
end
for x=1:size(z,2)
%test K-NN konvensional
cp=knnclassify(test, train, trainc, z(x), 'euclidean');
AB{x,1} = classperf(testc, cp); %akurasi acd(x)=AB{x,1}.CorrectRate;
%---sort---for n=1:size(dist,1)
[B,index]=sort(dist{n,1});
%index=index(1:z(x),1);
lmc=(sumc./ic);
distlma=sqrt(sum((test(n,:)-lma).^2));
distlmb=sqrt(sum((test(n,:)-lmb).^2));
distlmc=sqrt(sum((test(n,:)-lmc).^2));
distlha=1/wa;
distlhb=1/wb;
distlhc=1/wc;
if(ia~=0 && ib~=0 && ic~=0)
if(distlma<distlmb && distlma<distlmc) class(n,1)=1;
elseif (distlmb<distlma && distlmb<distlmc) class(n,1)=2;
else
class(n,1)=3;
end
if(distlha<distlhb && distlha<distlhc) class1(n,1)=1;
elseif(distlhb<distlha && distlhb<distlhc) class1(n,1)=2;