Rachimawan, Achmad Fachrudin (2016) ADS Filtering Menggunakan Jaringansyaraf Tiruan Perceptron, Naïve Bayes Classifier Dan Regresi Logistik. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
1311100123-Undergraduate Thesis.pdf - Published Version Download (2MB) | Preview |
Abstract
Email merupakan fasilitas yang mutlak diperlukan dalam berbagai bidang. Pentingnya email dan jumlahnya yang begitu banyak menyebabkan penyalahgunaan. Salah satu penyalahgunaan yang sering ditemui adalah email iklan yang dikirimkan oleh perusahaan penyedia konten internet saat pengguna mendaftar pada situs perusahaan tersebut. Terdapat metode agar email iklan dari perusahaan-perusahaan tersebut bisa secara otomatis dikenali yaitu klasifikasi. Data email berbentuk teks, sehingga jauh lebih rumit dan perlu proses untuk mempersiapkan data. Salah satu prosesnya adalah pembobotan ads atau adicity. Metode klasifikasi yang digunakan adalah Naive Bayes Classifier (NBC) yang secara umum sering digunakan dalam data teks dan Perceptron yang diketahui merupakan metode yang cukup sederhana untuk menyelesaikan permasalahan kompleks. Kedua metode tersebut akan dibandingkan dengan metode regresi logistik untuk mengetahui akurasi paling baik. Hasil penelitian menunjukkan bahwa NBC lebih unggul dibanding Perceptron dan Regresi Logistik, dan pada NBC False Positive Ratio lebih mudah untuk dikontrol.
=====================================================================================================
Email is a facility that is absolutely necessary in various fields. The importance of email its huge numbers causes of abuse. One of them is the advertisement emails sent by the provider of Internet content when users register on their company's website. There are several methods to solve that problem, one of them that advertisement email from these companies can automatically recognize is classification. Data on email is a text form, so it's much more complicated and need a different process to prepare the data. One of the process is the weighting of ads or adicity. The classification method used in this research is Naive Bayes classifier (NBC), which is often used in text data and Perceptron that known both of them which are fairly simple method to solve complex problems. Both of these methods will be compared with logistic regression to determine the best results. The results showed that the NBC superior to Perceptron and Logistic Regression , and on NBC False Positive Ratio is easier to control.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSt 006.3 Rac a |
| Uncontrolled Keywords: | email, iklan, klasifikasi, naïve bayes, perceptron |
| Subjects: | Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Mr. Tondo Indra Nyata |
| Date Deposited: | 23 Feb 2018 03:41 |
| Last Modified: | 27 Dec 2018 06:41 |
| URI: | http://repository.its.ac.id/id/eprint/51346 |
Actions (login required)
 |
View Item |