Aprianti, Winda (2015) Penerapan Rough Set Dan Fuzzy Rough Set Untuk Klasifikasi Data Tidak Lengkap. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
1213201029-Presentation-1213201029-presentationpdf.pdf - Published Version Download (1MB) | Preview |
Preview |
Text
1213201029-Master_Theses.pdf - Published Version Download (6MB) | Preview |
Abstract
Data mining merupakan salah satu proses untuk menemukan pola dan
pengetahuan dari database. Sebagian besar database di dunia nyata tidak dapat
dihindari dari masalah ketidaklengkapan. Hal ini disebabkan antara lain oleh
kesalahan prosedur manual entri data, pengukuran yang salah, dan kesalahan
peralatan. Salah satu database yang tidak terlepas dari masalah ketidaklengkapan
adalah dataset meteorologi, sehingga diperlukan algoritma klasifikasi yang
mampu menangani nilai atribut yang tidak lengkap dalam data meteorologi.
Dataset meteorologi yang digunakan terdiri dari atribut temperatur, kelembaban,
tekanan udara, kecepatan angin, dan curah hujan. Pada penelitian ini, penanganan
data yang tidak lengkap menggunakan algoritma klasifikasi berbasis rough set
dan fuzzy rough set. Hasil yang diperoleh berupa rules untuk mengklasifikasikan
data meteorologi tidak lengkap pada data uji.
Hasil pengujian algoritma rough set dan fuzzy rough set pada dataset yang
memuat 5%, 10%, 15%, 20%, 25%, dan 30% missing value menunjukkan bahwa:
(i) akurasi rules berbasis algoritma rough set mengalami penurunan ketika
persentase missing value bertambah, sedangkan akurasi rules berbasis algoritma
fuzzy rough set mengalami peningkatan ketika persentase missing value
ditingkatkan sampai 25% dan akurasi mengalami penurunan ketika persentase
missing value bertambah menjadi 30%, (ii) peningkatan persentase missing value
mempengaruhi jumlah rules dan waktu komputasi pembentukan rules berbasis
algoritma rough set, tetapi tidak berpengaruh pada jumlah rules dan waktu
komputasi pembentukan rules berbasis algoritma fuzzy rough set, dan (iii) pada
penerapan rules terhadap data uji, terdapat data uji yang tidak dapat diprediksi
oleh rules berbasis algoritma rough set, tetapi dapat diprediksi oleh rules berbasis
algoritma fuzzy rough set. ========== Data mining is a process of finding patterns and knowledge of the
database. Most of the databases in the real world can not be avoided from the
problem of incompleteness. This is caused partly by faulty procedure manual data
entry, wrong measurements, and equipment faults. One of database that can not be
separated from the problem of incompleteness is the meteorological dataset, so
that required classification algorithm that capable of handling incomplete attribute
values in meteorological data. Meteorological dataset is used consist of the
average temperature, humidity, air pressure, wind of speed, and rainfall. In this
study, the handling of incomplete data use classification algorithm based on rough
sets and fuzzy rough sets. The results obtained in the form of rules for classifying
the incomplete meteorological data on test data.
Results of the testing rough set and fuzzy rough set algorithm on
meteorological dataset containing 5%, 10%, 15%, 20%, 25%, and 30% missing
value showed that: (i) the accuracy of the rules based rough set algorithm
decreased when the percentage of missing value increases, while the accuracy of
the rules based fuzzy rough set algorithm increased when the percentage of
missing value increased to 25% and accuracy decreased when the percentage of
missing value increased to 30%, (ii) an increase in the percentage of missing value
affects the number of rules and computing time of forming rules based rough sets
algorithm, but it had no effect on the number of rules and computing time of
forming rules based fuzzy rough sets algorithms, and (iii) application of rules for
the test data, there is a test data that can not be predicted by the rules based rough
sets algorithm, but can be predicted by rules based fuzzy rough sets algorithms.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | klasifikasi; data tidak lengkap; rough set; fuzzy rough set; Classification; incomplete data; fuzzy rough set; rough set. |
| Subjects: | Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science) Q Science > QA Mathematics > QA9.58 Algorithms T Technology > T Technology (General) > T57.5 Data Processing |
| Divisions: | Faculty of Mathematics and Science > Mathematics > 44101-(S2) Master Thesis |
| Depositing User: | - Davi Wah |
| Date Deposited: | 24 May 2018 03:43 |
| Last Modified: | 24 May 2018 03:59 |
| URI: | http://repository.its.ac.id/id/eprint/51897 |
Actions (login required)
 |
View Item |