Putri, Rizky Amalia (2020) Prediksi Indeks Batubara di PT XYZ Menggunakan Ridge Regression dan Support Vector Regression (SVR). Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211640000040-Undergraduate_Thesis.pdf Download (2MB) | Preview |
Abstract
Semen merupakan salah satu bahan baku yang amat penting dalam pembangunan infrastruktur. Salah satu perusahaan yang bergerak di bidang produksi semen adalah PT XYZ. Tahapan terpenting dalam proses pembuatan semen adalah pada tahap pembakaran batu kapur dan tanah liat (clinker). Dalam proses pembakaran clinker membutuhkan ba-han bakar utama yaitu batubara. Semakin banyak jumlah produksi clin-ker yang dihasilkan dan semakin sedikit batubara yang digunakan dalam proses pembakaran, maka semakin efektif dan efisien proses produksi ter-sebut. Dalam penelitian ini dilakukan analisis untuk memprediksi indeks batubara dengan beberapa variabel yang diduga mempengaruhi yaitu kualitas batubara, bahan baku, dan operasional yang kemudian dilaku-kan estimasi terhadap indeks batubara. Metode yang digunakan untuk mengestimasi indeks batubara adalah metode Regresi Ridge dan metode Support Vector Regression (SVR). Sebelum dilakukan SVR, perlu dilaku-kan pemilihan variabel atau Feature Selection dengan metode Recursive Feature Elimination berbasis Random Forest. Model yang terbentuk de-ngan metode SVR dibandingkan dengan metode regresi ridge, kemudian dipilih model terbaiknya diantara kedua model yang terbentuk meng-gunakan nilai RMSE. Hasil analisis didapatkan metode terbaik dengan nilai RMSE terkecil yaitu Support Vector Regression (SVR) dengan com-plete feature dan menggunakan kernel-polynomial yang menghasilkan parameter sigma bernilai 0,100 dan nilai c sebesar 1 dengan nilai RMSE sebesar 0,619.
=================================================================================================================================
Cement is one of the most important raw materials in infrastruc-ture development. One of the companies engaged in cement production is PT XYZ. The most important stage in the process of making cement is at the stage of burning limestone and clay (clinker). In the process of burn-ing clinker requires the main fuel, namely coal. The more clinker produc-tion is produced and the less coal used in the combustion process, the mo-re effective and efficient the production process. In this study analysis will be conducted to predict the coal indeks with several variables that are thought to affect the quality of coal, raw materials, and operations which will then be estimated on the coal indeks. The method used to estimate the coal indeks is Ridge Regression method and Support Vector Regression (SVR) method. Prior to the SVR, a variable or Feature Selection will be conducted using the Random Forest-based Recursive Feature Elimina-tion method. The model formed by the SVR method will be compared with the ridge regression method which will then be chosen the best model bet-ween the two models formed using the RMSE value. The analysis results obtained the best method with the smallest RMSE value, is Support Vector Regression (SVR) with complete features and using kernel-polynomial that produces a sigma parameter of 0.100 and a c value of 1 with an RMSE value of 0.619.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSSt 519.535 Put p-1 2020 |
| Uncontrolled Keywords: | Batubara, Clinker, Indeks, Ridge, Support Vector Regression |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | RIZKY AMALIA PUTRI |
| Date Deposited: | 29 Apr 2024 03:08 |
| Last Modified: | 29 Apr 2024 03:08 |
| URI: | http://repository.its.ac.id/id/eprint/73681 |
Actions (login required)
 |
View Item |