Puspitasari, Diantika (2013) Implementasi Fuzzy Iterative Dichotomiser 3 (Fid3) Dalam Klasifikasi Tebu. Sarjana thesis, Universitas Brawijaya.
Abstract
Gula merupakan salah satu kebutuhan pokok masyarakat yang relatif murah. Tebu adalah tanaman bahan baku dari gula. Salah satu hal penting yang perlu diperhatikan dalam pengembangan usaha industri gula berbasis tebu adalah pemilihan dan komposisi varietas bibit unggul bermutu. Dengan varietas bibit unggul tebubermutu maka akan menghasilkan gula yang bermutu pula. Skripsi ini membahas penerapan algoritma Fuzzy Iterative Dichotomiser 3 (FID3) dalam klasifikasi tebu. Aturan yang dihasilkan dari metode Fuzzy ID3 akan digunakan untuk mendapatkan tebu yang bermutu unggulan. Algoritma Fuzzy ID3 adalah algoritma decision tree yang bekerja dengan cara membangun tree berdasarkan data latih, kemudian tree tersebut dikonversi menjadi aturan. Aturan yang terbentukakan digunakan sebagai dasar pada proses pengujian. Hasil dari pengujian ini selanjutnya digunakan untuk menentukan kelas suatu tebu dan menghitung akurasi sistem. Pengujian dilakukan terhadap aturan yang telah terbentuk pada proses pelatihan dengan berdasarkan kombinasi nilai Fuzziness Control Threshold (FCT) dan Leaf Decision Threshold (LDT). Nilai FCT yang digunakan 40%,45%,50%,60%,65%,70%,75%,80%,85%,90%,95% dan 98%. Sedangkan nilai LDT yang digunakan adalah 3%,5%,8%,10%,15% dan 18%. Pengujian untuk mendapatkan aturan yang terbaik dilakukan terhadap 4 macam data latih yaitu 100, 200, 260, dan 300.Pengujian akurasi dilakukan pada jumlah data uji yang sama untuk setiap data latihnya, yaitu 50 data. Akurasi tertinggi dari pembentukan aturan akan diuji sebagai aturan terbaik. Hasil pengujian aturan terbaik yang dilakukan menunjukkan jumlah aturan terbaik berasal dari 300 data latih dengan akurasi 96% dan jumlah aturan 100 aturan pada nilai FCT 45% serta nilai LDT 3%.Penelitian ini menunjukkan bahwa algoritma Fuzzy Iterative Dichotomiser 3 (FID3) memiliki kinerja yang baik dalam klasifikasi tebu.
English Abstract
Sugar is one of this nation strategical comodity where it is a priority needs of people, and also sugar is low price calory resources. Canes are plants that only grows in tropical land and canes is raw materials of sugar. One of important thing that should be noticed in sugar industry development especially cane based is selection process and composition of good quality cane seeds because good quality sugar come from good quality cane. This research discuss about implementation of Fuzzy Iterative Dichotomiser 3 (FID3) algorithm for sugar classification. Fuzzy ID3 is algorithm based on decision tree that generate classification tree based on training data. The rules will be composed based on that tree that traversed from root until its lowest leaf. These rules will be used to determine class of the cane and to calculate the accuration of system. This algorithm will be evaluated by analyzing rules generated in learning process. The rule also analyzed by combination of Fuzziness Control Threshold (FCT) and Leaf Decision Threshold (LDT). FCT values that used are 40%, 45%, 50%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95% and 98% while LDT value that used are 3%, 5%, 8%, 10%, 15% and 18%. Experimental test that will be done to find best rule will use various number of data which are 100, 200, 260 and 300. Experimental scenario for this research to find accuracy of the system is the number of test data will be constant which is 50 while the number of training data is changeable. Experimental results show the highest number of rule that is generated by the system is 100 rules with FCT and LDT value respectively 45% and 3% and the accuracy value is 96%. This result show that Fuzzy Iterative Dichotomiser 3 (FID3) algorithm has good performance for cane classification.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/FTIK/2013/173/051311815 |
| Subjects: | 000 Computer science, information and general works > 005 Computer programming, programs, data |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Hasbi |
| Date Deposited: | 24 Dec 2013 14:32 |
| Last Modified: | 23 Oct 2021 09:38 |
| URI: | http://repository.ub.ac.id/id/eprint/145802 |
Preview |
Text
5-KATA_PENGANTAR_(iv).pdf Download (1MB) | Preview |
Preview |
Text
1-cover.pdf Download (1MB) | Preview |
Preview |
Text
3-LEMBAR_PENGESAHAN_(ii).pdf Download (1MB) | Preview |
Preview |
Text
6-ABSTRAK(vi).pdf Download (1MB) | Preview |
Preview |
Text
4-PERNYATAAN_ORISINALITAS_SKRIPSI_(iii).pdf Download (1MB) | Preview |
Preview |
Text
2-Lembar_Persetujuan_(i).pdf Download (1MB) | Preview |
Preview |
Text
7-Daftar_Isi_(viii).pdf Download (1MB) | Preview |
Preview |
Text
8-Daftar_Gambar(xii).pdf Download (1MB) | Preview |
Preview |
Text
10-Daftar_Source_code(xvi).pdf Download (1MB) | Preview |
Preview |
Text
9-Daftar_Tabel(xiv).pdf Download (1MB) | Preview |
Preview |
Text
BAB_I_(1-4).pdf Download (1MB) | Preview |
Preview |
Text
BAB_II_(5-28).pdf Download (2MB) | Preview |
Preview |
Text
BAB_III_(29-84).pdf Download (4MB) | Preview |
Preview |
Text
BAB_V_(110-121).pdf Download (2MB) | Preview |
Preview |
Text
BAB_IV_(85-109).pdf Download (3MB) | Preview |
Preview |
Text
BAB_VI_(122).pdf Download (1MB) | Preview |
Preview |
Text
DAFTAR_PUSTAKA_(123-124).pdf Download (1MB) | Preview |
Preview |
Text
Lampiran_(125-144).pdf Download (4MB) | Preview |
Actions (login required)
 |
View Item |