SEGMENTASI PELANGGAN MENGGUNAKAN METODE PARTICLE SWARM OPTIMIZATION DAN K-MEANS TUGAS AKHIR Diajukan guna memenuhi sebagian persyaratan dalam rangka menyelesaikan Pendidikan Sarjana Strata Satu (S1) Program Studi Teknologi Informasi I DEWA AYU AGUNG YUNITA PRIMANDARI NIM: 1204505063 JURUSAN TEKNOLOGI INFORMASI FAKULTAS TEKNIK UNIVERSITAS UDAYANA 2016
14
Embed
SEGMENTASI PELANGGAN MENGGUNAKAN … Awal.pdf · 4.2.5 Proses Optimasi Titik Pusat Cluster dengan Metode Particle Swarm Optimization ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
SEGMENTASI PELANGGAN MENGGUNAKAN
METODE PARTICLE SWARM OPTIMIZATION
DAN K-MEANS
TUGAS AKHIR
Diajukan guna memenuhi sebagian persyaratan dalam rangka menyelesaikan
Pendidikan Sarjana Strata Satu (S1) Program Studi Teknologi Informasi
I DEWA AYU AGUNG YUNITA PRIMANDARI
NIM: 1204505063
JURUSAN TEKNOLOGI INFORMASI
FAKULTAS TEKNIK
UNIVERSITAS UDAYANA
2016
ii
PERNYATAAN
Dengan ini saya menyatakan bahwa dalam Tugas Akhir ini tidak terdapat
karya yang pernah diajukan untuk memperoleh gelar kesarjanaan di perguruan
tinggi lain, dan sepanjang pengetahuan saya tidak terdapat karya atau pendapat
yang pernah ditulis atau diterbitkan oleh orang lain, kecuali yang secara tertulis
diacu dalam naskah ini dan disebutkan pada daftar pustaka.
Denpasar, Juni 2016
I Dewa Ayu Agung Yunita Primandari
iii
LEMBAR PENGESAHAN TUGAS AKHIR
iv
BERITA ACARA TUGAS AKHIR
v
KATA PENGANTAR
Puji dan syukur penulis panjatkan kehadapan Ida Sang Hyang Widhi
Wasa/Tuhan Yang Maha Esa, karena atas Asung Kerta Wara Nugraha-Nya,
akhirnya penulis dapat menyelesaikan tugas akhir dengan judul “Segmentasi
Pelanggan Menggunakan Metode Particle Swarm Optimization dan K-Means”.
Penulis mendapat banyak bimbingan dari berbagai pihak. Ucapan terima kasih
penulis sampaikan kepada:
1. Bapak Prof. Ir. Ngakan Putu Gede Suardana, M.T., Ph.D. selaku Dekan
Fakultas Teknik Universitas Udayana.
2. Bapak Dr. Eng. I Putu Agung Bayupati, ST., MT, selaku Ketua Jurusan
Teknologi Informasi Universitas Udayana.
3. Bapak Prof. Dr. I Ketut Gede Darma Putra, S.Kom., M.T., selaku
dosen pembimbing I yang telah banyak memberikan bimbingan dan masukan
dalam penyusunan tugas akhir ini.
4. Bapak I Made Sukarsa, S.T., M.T., selaku dosen pembimbing II yang telah
banyak memberikan petunjuk dan bimbingan selama penyusunan tugas akhir.
HALAMAN DEPAN ............................................................................................. i
PERNYATAAN ..................................................................................................... ii
LEMBAR PENGESAHAN TUGAS AKHIR .................................................... iii BERITA ACARA TUGAS AKHIR ................................................................... iv KATA PENGANTAR ........................................................................................... v ABSTRAK .......................................................................................................... vi
ABSTRACT ......................................................................................................... vii DAFTAR ISI ....................................................................................................... viii DAFTAR GAMBAR ............................................................................................. x DAFTAR TABEL ............................................................................................... xii
DAFTAR KODE PROGRAM .......................................................................... xiv BAB I PENDAHULUAN ..................................................................................... 1 1.1 Latar Belakang ............................................................................................... 1 1.2 Rumusan Masalah .......................................................................................... 3
1.3 Tujuan Penelitian ........................................................................................... 3 1.4 Manfaat Penelitian ......................................................................................... 3 1.5 Batasan Masalah ............................................................................................ 4
BAB II KAJIAN PUSTAKA ............................................................................... 6 2.1 State of the Art ............................................................................................... 6 2.2 Data Mining ................................................................................................. 10
2.3 Hubungan Data Mining dalam Kerangka Kerja CRM ................................ 11 2.4 Model RFM ................................................................................................. 13
2.5 Normalisasi Data ......................................................................................... 15 2.6 Metode Clustering ....................................................................................... 16 2.7 Metode K-Means ......................................................................................... 16 2.8 Metode Particle Swarm Optimization ......................................................... 19
2.8.1 Personal Best dan Global Best Particle Swarm Optimization ......... 20 2.8.3 Contoh Perhitungan Particle Swarm Optimization .......................... 21
2.9 Validasi Cluster ........................................................................................... 23 2.9.1 Validasi Cluster dengan Davies-bouldin Index ............................... 23
2.9.2 Validasi Cluster dengan Silhouette Index ........................................ 24 2.10 Profil Perusahaan PT. X .............................................................................. 25
BAB III METODOLOGI PENELITIAN ........................................................ 26 3.1 Tempat dan Waktu Penelitian ...................................................................... 26 3.2 Alur Penelitian ............................................................................................. 26 3.3 Sumber Data ................................................................................................ 27 3.4 Kebutuhan Hardware dan Software ............................................................ 27 3.5 Perancangan Sistem ..................................................................................... 28
3.5.1 Gambaran Umum Sistem ................................................................. 28 3.5.2 Pemilihan Data ................................................................................. 30
3.5.4 Transformasi Data ............................................................................ 35 3.5.5 Normalisasi Data .............................................................................. 39
ix
3.5.6 Optimasi Titik Pusat Cluster Menggunakan PSO ........................... 44 3.5.7 Clustering Menggunakan K-Means ................................................. 51
3.5.8 Pemodelan Data ............................................................................... 58 3.5.9 Cluster Validation dengan Davies-bouldin Index ........................... 61 3.5.10 Cluster Validation dengan Silhouette Index .................................... 66
3.6 Perancangan Basis Data ............................................................................... 69 3.7 Interface ....................................................................................................... 79
BAB IV PEMBAHASAN DAN ANALISA HASIL ........................................ 85 4.1 Piranti Implementasi Aplikasi ..................................................................... 85
4.1.1 Kebutuhan Perangkat Lunak Implementasi Aplikasi ...................... 85 4.1.2 Kebutuhan Perangkat Keras Implementasi Aplikasi ....................... 85
4.2 Hasil Perancangan Aplikasi ......................................................................... 86 4.2.1 Tampilan Menu ................................................................................ 86 4.2.2 Proses Pemilihan Data ..................................................................... 87
4.2.3 Proses Transformasi Data ................................................................ 88 4.2.4 Proses Normalisasi Data .................................................................. 90
4.2.5 Proses Optimasi Titik Pusat Cluster dengan Metode Particle Swarm
4.2.6 Proses Clustering Menggunakan Metode K-Means ........................ 92 4.2.7 Proses Validasi Cluster .................................................................... 93
4.3.1 Hasil Uji Coba Clustering dengan 2 Cluster ................................... 95
4.3.2 Hasil Uji Coba Clustering dengan 3 Cluster ................................... 98 4.3.3 Hasil Uji Coba Clustering dengan 4 Cluster ................................. 102 4.3.4 Hasil Uji Coba Clustering dengan 5 Cluster ................................. 106
4.3.5 Hasil Uji Coba Clustering dengan 6 Cluster ................................. 110 4.4 Analisis Cluster ......................................................................................... 114
4.5 Perbandingan Hasil Clustering Metode K-Means dan Metode PSO+K-
Means ......................................................................................................... 115 4.6 Perbandingan Kelas Konsumen per Bulan ................................................ 119
BAB V PENUTUP ............................................................................................ 123
DAFTAR PUSTAKA ........................................................................................ 125
x
DAFTAR GAMBAR
Gambar 2.1 Diagram Fishbone............................................................................... 9 Gambar 2.2 Proses KDD ....................................................................................... 10 Gambar 2.3 Skema CRIPS-DM ............................................................................ 12
Gambar 2.4 Topologi Bintang............................................................................... 20 Gambar 3.1 Gambaran Umum Sistem .................................................................. 28 Gambar 3.2 Relasi Tabel Data Transaksi PT. X ................................................... 31
Gambar 3.3 Skema Pembentukan Nilai RFM ....................................................... 35
Gambar 3.4 Skema Penentuan Nilai tb_standar ............................................ 37
Gambar 3.5 Skema Penentuan Nilai tb_rfm ...................................................... 38
Gambar 3.6 Screenshoot Data tb_rfm ............................................................... 40
Gambar 3.7 Screenshoot Hasil Normalisasi Data tb_rfm .................................. 44
Gambar 3.8 Flowchart Algoritma PSO ................................................................ 46
Gambar 3.9 Hasil Iterasi Pertama Metode PSO .................................................... 49
Gambar 3.10 Hasil Iterasi Kedua Metode PSO .................................................... 51 Gambar 3.11 Flowchart Algoritma K-Means ....................................................... 52 Gambar 3.12 Hasil Iterasi Pertama Metode K-Means .......................................... 55
Gambar 3.13 Hasil Iterasi Kedua Metode K-Means ............................................. 57
Gambar 3.14 Flowchart Algoritma Davies-bouldin Index ................................... 62 Gambar 3.15 Flowchart Algoritma Silhouette ...................................................... 67 Gambar 3.16 Rancangan Basis Data Bantu .......................................................... 70
Gambar 3.17 Form Home ..................................................................................... 79 Gambar 3.18 Form Transfomasi Monthly Segmentation ...................................... 80
Gambar 3.19 Form Transformasi Data ................................................................. 80 Gambar 3.20 Form Pemilihan Titik Pusat ............................................................ 81 Gambar 3.21 Form Clustering .............................................................................. 81
Gambar 3.22 Form Cluster Validation ................................................................. 82 Gambar 3.23 Form Hasil Segmentasi ................................................................... 83 Gambar 3.24 Form Hasil Segmentasi Per Pelanggan ........................................... 83
Gambar 3.25 Form About ..................................................................................... 84
Gambar 4.1 Tampilan Menu ................................................................................. 86
Gambar 4.2 Proses Pemilihan Data ....................................................................... 87
Gambar 4.3 Skema Pembentukan tb_standar ................................................ 88
Gambar 4.4 Skema Transformasi Data ................................................................. 88
Gambar 4.5 Proses Normalisasi Data ................................................................... 90 Gambar 4.6 Proses Optimasi Titik Pusat Cluster ................................................. 92 Gambar 4.7 Proses Clustering dengan Metode K-Means ..................................... 93 Gambar 4.8 Proses Validasi Cluster ..................................................................... 94 Gambar 4.9 Proses Pemilihan Titik Pusat dengan 2 Cluster ................................ 95
Gambar 4.10 Titik Pusat Optimal dengan 2 Cluster ............................................. 96 Gambar 4.11 Hasil Clustering K-Means dengan 2 Cluster ................................... 96
Gambar 4.12 Hasil Segmentasi Pembentukan 2 Cluster ...................................... 97 Gambar 4.13 Proses Pemilihan Titik Pusat dengan 3 Cluster .............................. 98 Gambar 4.14 Titik Pusat Optimal dengan 3 Cluster ............................................. 99
xi
Gambar 4.15 Hasil Clustering K-Means dengan 3 Cluster ................................. 100 Gambar 4.16 Hasil Segmentasi Pembentukan 3 Cluster .................................... 101
Gambar 4.17 Proses Pemilihan Titik Pusat dengan 4 Cluster ............................ 102 Gambar 4.18 Titik Pusat Optimal dengan 4 Cluster ........................................... 103 Gambar 4.19 Hasil Clustering K-Means dengan 4 Cluster ................................. 104 Gambar 4.20 Hasil Segmentasi Pembentukan 4 Cluster .................................... 105 Gambar 4.21 Proses Pemilihan Titik Pusat dengan 5 Cluster ............................ 106
Gambar 4.22 Titik Pusat Optimal dengan 5 Cluster ........................................... 107 Gambar 4.23 Hasil Clustering K-Means dengan 5 Cluster ................................. 108
Gambar 4.24 Grafik Clustering K-Means dengan 5 Cluster .............................. 109 Gambar 4.25 Proses Pemilihan Titik Pusat dengan 6 Cluster ............................ 110 Gambar 4.26 Titik Pusat Optimal dengan 6 Cluster ........................................... 111 Gambar 4.29 Hasil Clustering K-Means dengan 6 Cluster ................................. 112 Gambar 4.28 Grafik Clustering K-Means dengan 6 Cluster .............................. 113
Gambar 4.29 Hasil Validasi Cluster ................................................................... 115 Gambar 4.30 Perbandingan Hasil Cluster Metode K-Means dan Metode
PSO+K-Means ............................................................................... 116 Gambar 4.30 Grafik Segementasi Pelanggan (Kode Pelanggan = 12100) ......... 120
Gambar 4.31 Grafik Segementasi Pelanggan (Kode Pelanggan = 11005) ......... 121 Gambar 4.32 Grafik Segementasi Pelanggan (Kode Pelanggan = 12493) ......... 121 Gambar 4.33 Grafik Segementasi Pelanggan (Kode Pelanggan = 11794) ......... 122
xii
DAFTAR TABEL
Tabel 2.1 Daftar State of the Art ............................................................................. 8 Tabel 2.2 Klasifikasi Customer ............................................................................ 14
Tabel 2.3 Data Sumber .......................................................................................... 17 Tabel 2.4 Titik Pusat Iterasi ke-1 .......................................................................... 18 Tabel 2.5 Hasil Perhitungan Jarak ........................................................................ 19
Tabel 3.10 Contoh Tabel Detail Transaksi ........................................................... 35 Tabel 3.11 Data Pemilihan Atribut Sesuai Model RFM ....................................... 35 Tabel 3.12 Contoh Data pada Tabel Transaksi ..................................................... 36
Tabel 3.13 Contoh Data pada tb_standar ....................................................... 37
Tabel 3.14 Contoh Data pada tb_rfm ................................................................ 39
Tabel 3.15 Contoh Data pada tb_rfm ................................................................ 39
Tabel 3.16 Contoh Data Atribut Recency ............................................................. 40
Tabel 3.17 Contoh Data Atribut Frequency .......................................................... 42 Tabel 3.18 Contoh Data Atribut Monetary ........................................................... 43
Tabel 3.19 Hasil Normalisasi Data tb_rfm ........................................................ 44
Tabel 3.21 Normalisasi Data tb_rfm ................................................................. 47
Tabel 3.21 Contoh Data pada tb_rfm yang Ternormalisasi ............................. 52
Tabel 3.22 Perhitungan Jarak Iterasi Pertama ....................................................... 54 Tabel 3.23 Perhitungan Jarak Iterasi Pertama ....................................................... 57
Tabel 3.24 Domain Nilai RFM ............................................................................. 58 Tabel 3.25 Deskripsi Variabel Linguistik dan Label Konsumen .......................... 59
Tabel 3.26 Data tb_rfm ...................................................................................... 60
Tabel 3.27 Domain Nilai RFM ............................................................................. 61 Tabel 3.28 Kelas Cluster ....................................................................................... 61
Tabel 3.29 Normalisasi Data tb_rfm ................................................................. 63
Tabel 3.30 Contoh Data Atribut Cluster 1 ............................................................ 63 Tabel 3.31 Contoh Data Atribut Cluster 2 ............................................................ 64
Tabel 3.32 Normalisasi Data tb_rfm ................................................................. 68
Tabel 3.33 Hasil Perhitungan Indeks Silhouette ................................................... 69
Tabel 3.34 Struktur tb_standar ...................................................................... 70
Tabel 3.35 Contoh Data tb_standar ............................................................... 71
Tabel 3.36 Struktur tb_rfm ................................................................................ 71
xiii
Tabel 3.37 Struktur Data tb_rfm ........................................................................ 72
Tabel 3.38 Struktur tb_hasil ........................................................................... 72
Tabel 3.39 Struktur Data tb_hasil ................................................................... 72
Tabel 3.40 Struktur tb_range_r ...................................................................... 73
Tabel 3.41 Struktur Data tb_hasil_r .............................................................. 73
Tabel 3.42 Struktur tb_range_f ...................................................................... 74
Tabel 3.43 Struktur Data tb_hasil_f .............................................................. 74
Tabel 3.44 Struktur tb_range_m ...................................................................... 74
Tabel 3.45 Struktur Data tb_hasil_monetary ............................................. 75
Tabel 3.46 Struktur tb_kelas ........................................................................... 75
Tabel 3.47 Struktur Data tb_kelas ................................................................... 76
Tabel 3.48 Struktur tb_avg ................................................................................ 76
Tabel 3.49 Struktur Data tb_avg ........................................................................ 77
Tabel 3.50 Struktur tb_cluster ...................................................................... 77
Tabel 3.51 Struktur Data tb_cluster .............................................................. 77
Tabel 3.52 Struktur tb_hasil_normal .......................................................... 78
Tabel 3.53 Struktur Data tb_normal ................................................................ 78
Tabel 3.54 Struktur tb_optimasi .................................................................... 78
Tabel 3.57 Struktur Data tb_optimasi ............................................................ 79
Tabel 4.1 Hasil Optimasi dengan Parameter Cluster =2 dan Iterasi = 20 ............. 95 Tabel 4.2 Hasil Segmentasi Pembentukan 2 Cluster ............................................ 97
Tabel 4.3 Hasil Segmentasi Pembentukan 2 Cluster ............................................ 97 Tabel 4.4 Hasil Optimasi dengan Parameter Cluster =3 dan Iterasi = 20 ............. 99 Tabel 4.5 Hasil Segmentasi Pembentukan 3 Cluster .......................................... 100
Tabel 4.6 Hasil Segmentasi Pembentukan 3 Cluster .......................................... 100 Tabel 4.7 Hasil Optimasi dengan Parameter Cluster =4 dan Iterasi = 20 ........... 103
Tabel 4.8 Hasil Segmentasi Pembentukan 4 Cluster .......................................... 104 Tabel 4.9 Hasil Segmentasi Pembentukan 4 Cluster .......................................... 104
Tabel 4.10 Hasil Optimasi dengan Parameter Cluster = 5 dan Iterasi = 20 ........ 107 Tabel 4.11 Hasil Segmentasi Pembentukan 5 Cluster ........................................ 108 Tabel 4.12 Hasil Segmentasi Pembentukan 5 Cluster ........................................ 108
Tabel 4.13 Hasil Optimasi dengan Parameter Cluster = 6 dan Iterasi = 20 ........ 111 Tabel 4.14 Hasil Segmentasi Pembentukan 6 Cluster ........................................ 112 Tabel 4.15 Hasil Segmentasi Pembentukan 6 Cluster ........................................ 113 Tabel 4.16 Hasil Segmentasi Pembentukan 6 Cluster ........................................ 117
Tabel 4.17 Waktu Eksekusi Metode ................................................................... 118 Tabel 4.16 Perbandingan Kelas Pelanggan per Bulan ........................................ 119
xiv
DAFTAR KODE PROGRAM
Kode Program 4.1 Stored Procedure Input Nilai RFM ........................................ 89
Kode Program 4.2 Fungsi Normalisasi Data dengan Metode Min and Max ........ 90 Kode Program 4.3 Fungsi Metode Particle Swarm Optimization ........................ 91 Kode Program 4.4 Fungsi Metode K-Means ........................................................ 93 Kode Program 4.5 Fungsi Metode DBI ................................................................ 93
Kode Program 4.6 Fungsi Metode Silhouette Index ............................................. 94