• Login
    View Item 
    •   USU-IR Home
    • Faculty of Computer Science and Information Technology
    • Department of Information Technology
    • Master Theses
    • View Item
    •   USU-IR Home
    • Faculty of Computer Science and Information Technology
    • Department of Information Technology
    • Master Theses
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Analisis Kinerja Metode Dbscan (Density-Based Spatial Clustering of Applications with Noise) dan K-Means dalam Sistem Pendukung Keputusan

    View/Open
    Fulltext (2.163Mb)
    Date
    2017
    Author
    Nur, Fauziah
    Advisor(s)
    Zarlis, Muhammad
    Nasution, Benny Benyamin
    Metadata
    Show full item record
    Abstract
    Clustering is one of the unsupervised mining data methods; it is also a method used to seek and to group data which have characteristic resemblance from one datum to another one. In this research, K-means method and DBSCAN method were used to group the data; rule-based classification was also used to the supplementary clustering method as a comparison between K-Means method and DBSCAN method. In this case, the research grouped 6 clusters at SMK Swasta (Private Vocational School) Medan Area 1 by using the criteria found in the students’ data such as sex, parents’ income, parents’ dependents, test scores, and students’ body height. There were 40 data of the students as the samples. The result of rule-based classification was 648 rules. The data were then grouped by using K-Means method which yielded 6 clusters with only 3 noises. Grouping occurred until 4 iterations. In DBSCAN method, the data were grouped by using 2 parameters: epsilon = 0.00972 and MinPts = 2 which yielded 3 clusters. After grouping by using both methods, the data were tested by using non-parametric statistical test; the result was Zcount = 4.8 so the Ho was rejected and Hi was accepted which indicated that using K-Means was more optimal than using DBSCAN in this research. This clustering was beneficial to group the students according some criteria which had determined and tested the performance of the used methods.
     
    Clustering merupakan salah satu metode data mining yang bersifat tanpa arahan (unsupervised) dan suatu metode untuk mencari dan mengelompokan data yang memiliki kemiripan karakteristik antara satu data dengan data lain. Dalam penelitian ini, untuk pengelompokan data menggunakan metode K-Means dan metode DBSCAN, adapun tambahan metode pengelompokan yaitu rule-based classification sebagai perbandingan antara metode K-Means dan DBSCAN. Dalam hal ini, peneliti mengelompokan 6 kelompok jurusan pada sekolah menengah kejuruan SMK Swasta Medan Area.1 dengan menggunakan kriteria – kriteria yang terdapat dalam data siswa tersebut seperti jenis kelamin, pendapatan orang tua, tanggungan anak orang tua, nilai tes dan tinggi badan siswa. Data sampel yang diuji adalah berjumlah 40 data siswa SMK. Pada penelitian ini, untuk pengujian rule-based classification menghasilkan 648 rule. Kemudian data dikelompokan dengan menggunakan k-Means yang menghasilkan 6 kelompok, dengan adanya 2 noise. Pengelompokan terjadi hingga 4 kali iterasi. Pada metode DBSCAN pengelompokan data menggunakan 2 parameter yaitu epsilon= 0.00972 dan MinPts= 2 yang menghasilkan 3 kelompok. Setelah hasil pengelompokan dengan menggunakan kedua metode tersebut, selanjutnya data diuji menggunakan uji statistik nonparametrik dengan hasil Zhitung = 4.8 sehingga didapat hasil menolak H0 dan menerima H1 yang berarti penggunaan K-Means lebih optimal daripada penggunaan DBSCAN dalam penelitian ini. Pengelompokan ini bermanfaat untuk mengelompokan siswa sesuai jurusan berdasarkan beberapa kriteria yang telah ditentukan dan menguji kinerja metode yang digunakan.

    URI
    http://repositori.usu.ac.id/handle/123456789/41667
    Collections
    • Master Theses [621]

    Repositori Institusi Universitas Sumatera Utara (RI-USU)
    Universitas Sumatera Utara | Perpustakaan | Resource Guide | Katalog Perpustakaan
    DSpace software copyright © 2002-2016  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    Atmire NV
     

     

    Browse

    All of USU-IRCommunities & CollectionsBy Issue DateTitlesAuthorsAdvisorsKeywordsTypesBy Submit DateThis CollectionBy Issue DateTitlesAuthorsAdvisorsKeywordsTypesBy Submit Date

    My Account

    LoginRegister

    Repositori Institusi Universitas Sumatera Utara (RI-USU)
    Universitas Sumatera Utara | Perpustakaan | Resource Guide | Katalog Perpustakaan
    DSpace software copyright © 2002-2016  DuraSpace
    Contact Us | Send Feedback
    Theme by 
    Atmire NV