Jurnal
Penerapan Sparse Principal Component Analysis dalam Menghasilkan Matriks Loading yang Sparse
Jurnal Matematika, Statistika dan Komputasi · Vol. 15 No. 2 (2018) · pp. 44–53
- ISSN
- 2614-8811, 1858-1382
- Published
- 2018-12-20
- Publisher
- Department of Mathematics, Hasanuddin University
- Language
- en
Abstract
Abstract Sparse Principal Component Analysis (Sparse PCA) is one of the development of PCA. Sparse PCA modifies new variables as a linier combination of p old variables (original variable) which is yielded by PCA method. Modifying new variables is conducted by producing a loading yang sparse matrix, such that old variable which is not effective (value of loading is zero) able be exit from PCA. In this study, Sparse PCA method was applied on data of Indonesia Poverty population in 2015, that contains 13 variables and 34 observation with variable reduction such that yields 4 (four) new variables, which can explain 80.1% of total variance data. This study show, the loading matrix that has been yielded by using Sparse PCA method to become sparse with there exist 11 elements (loading value) zero entry of matrix, such that the model that has been produced to be simpler and easy to be interpreted. Keywords: Principal Component Analysis, Sparse Principal Component Analysis, reduksi dimensi, matriks loading yang sparse Abstrak Sparse Principal Component Analysis (Sparse PCA) merupakan salah satu pengembangan dari metode PCA. Sparse PCA memodifikasi variabel-variabel baru yang merupakan kombinasi linear dari variabel lama (variabel asli) yang dihasilkan oleh metode PCA. Pemodifikasian variabel baru ini dilakukan dengan dengan menghasilkan matriks loading yang sparse sehingga variabel lama yang tidak efektif (memiliki nilai loading sama dengan nol) dapat dikeluarkan dari model PCA. Pada penelitian ini, metode Sparse PCA diterapkan pada data Indikator Kemiskinan Penduduk Indonesia Tahun 2015 yang memuat 13 variabel dan 34 observasi dengan reduksi variabel menghasilkan 4 (empat) variabel baru yang telah mampu menjelaskan 80,1% dari total variansi data. Hasil penelitian menunjukkan, matriks loading yang dihasilkan menggunakan metode Sparse PCA menjadi sparse dengan terdapat 11 elemen (nilai loading) matriks bernilai nol sehingga model yang dihasilkan menjadi lebih sederhana dan mudah untuk diinterpretasikan. Kata Kunci: Principal Component Analysis, Sparse Principal Component Analysis, reduksi dimensi, matriks loading yang sparse
Keywords
Collaboration network of the authors
35 authors · 113 co-authorship links
Hover or tap a name to highlight its co-authors. Double-click to open the author's page.
- Universitas Islam Negeri Sjech M. Djamil Djambek Bukittinggi
- Institut Agama Islam Negeri Kendari
- Jenderal Soedirman University
- Affiliation unknown
- Different organisations
- Same organisation
- Affiliation unknown
- Thicker line = more joint articles · bigger circle = more articles
Show as table
| Co-author | Organisation | Joint articles |
|---|---|---|
| Erna Tri Herdiani | Affiliation unknown | 3 |
| Andi Isna Yunita | Affiliation unknown | 3 |
| Siswanto Siswanto | Affiliation unknown | 2 |
| Nur Aminah Ahmad | Affiliation unknown | 2 |
| Raupong Raupong | Affiliation unknown | 1 |
| Vieri Koerniawan | Affiliation unknown | 1 |
| Anna Islamiyati | Affiliation unknown | 1 |
| Novilia Jao | Affiliation unknown | 1 |
| Riska Rasyid | Affiliation unknown | 1 |
| Andi Kresna Jaya | Affiliation unknown | 1 |
| Ika Pratiwi Haya | Affiliation unknown | 1 |
| - Siswanto | Affiliation unknown | 1 |
| Nirwan Ilyas | Affiliation unknown | 1 |
| Ayu Pratika Putri | Affiliation unknown | 1 |
| Annurial Fitrayah Taufiq | Affiliation unknown | 1 |
| Ahmad Maulana Syarif | Affiliation unknown | 1 |
| Nurmimie Kharina | Affiliation unknown | 1 |
| Samsir Aditya Ania | Affiliation unknown | 1 |
| Herlina Rasyid | Universitas Islam Negeri Sjech M. Djamil Djambek Bukittinggi | 1 |
| Nur Aliah Ramadhani | Affiliation unknown | 1 |
| Nadia Nazwadiah Caesar | Affiliation unknown | 1 |
| Zaraswati Dwyana Zainuddin | Affiliation unknown | 1 |
| Nurfadilah Nurfadilah | Institut Agama Islam Negeri Kendari | 1 |
| Andi M Alfin Baso | Affiliation unknown | 1 |
| Muhammad Ridzky Davala | Affiliation unknown | 1 |
| Agung Muhammad Takdir | Affiliation unknown | 1 |
| Nasrah Sirajang | Affiliation unknown | 1 |
| YULIANA.DEWI Putri | Affiliation unknown | 1 |
| Amran Amran | Jenderal Soedirman University | 1 |
| Hanifah Lainun | Affiliation unknown | 1 |
| Puji Puspa Sari | Affiliation unknown | 1 |
| Aidawayati Rangkuti | Affiliation unknown | 1 |
| Aprizal Resky | Affiliation unknown | 1 |
Affiliations come from OpenAlex; journal OAI-PMH metadata carries none. Authors without an OpenAlex ID are matched by name, which can merge or split people.