Febriasto, Rizki (2019) Prediksi Kuat Tekan Semen Untuk Produk Portland Composite Cement (PCC) di PT. Semen Indonesia (Persero) Tbk. Menggunakan Support Vector Regression (SVR) Dengan Feature Selection. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211745000020-Undergraduate_Theses.pdf - Published Version Download (1MB) | Preview |
Abstract
Indonesia sebagai negara berkembang terus-menerus melakukan pembangunan dalam segala bidang aspek kehidupan. Salah satu aspek yang selalu berkembang adalah sektor infrastruktur. Dengan menghadapi fenomena pertumbuhan infrastruktur ini, khususnya di sektor pembangunan perlu adanya hal yang menunjang dalam setiap pembangunan yang terjadi, salah satunya adalah material semen. Salah satu perusahaan yang bergerak di bidang produksi semen adalah PT. Semen Indonesia (Persero) Tbk. Terdapat beberapa produk yang dihasilkan oleh PT. Semen Indonesia (Persero) Tbk., salah satunya adalah Portland Composite Cement (PCC). Dalam penelitian ini akan dilakukan prediksi terhadap kuat tekan semen PCC di hari ke-28. Data yang digunakan yaitu bulan Juli 2018 hingga Maret 2019. Pada hasil analisis, hasil regresi linier telah terindikasi mengalami multikolineari-tas, sehingga ditangani menggunakan PCR. Feature selection yang digunakan adalah Recursive Feature Elimination (RFE) berbasis Random Forest (RF-RFE) yang telah menghilangkan lima variabel. Dengan teknik machine learning yaitu SVR didapatkan model terbaik yaitu SVR dengan fungsi kernel Radial Basis Function (RBF) dengan complete feature. Feature selection tidak menghasilkan hasil yang lebih baik dibanding complete feature. Model yang terbaik digunakan untuk memprediksi kuat tekan pada bulan April 2019 yang menghasilkan kriteria terbaik RMSE sebesar 8,78.
==============================================================================================================================
Indonesia as a developing country has continued to carry out development in all sectors of life. One of the sectors that are always developing is the infrastructure sector. By dealing with the phenomenon of infrastructure growth, especially in the development sector, it is necessary to have things to support in every development matter, one of them is cement material. One of the companies which engaged in cement production is PT. Semen Indonesia (Persero) Tbk. There are several products manufactured by PT. Semen Indonesia (Persero) Tbk. One of them is Portland Composite Cement (PCC). In this observation, predictions will be made on the compressive strength of PCC cement on the 28th day. The data used in this observation is in July 2018 until March 2019. The result of linear regression analysis is, it has been indicated to multicollinearity, then it is solved by PCR. Feature selection based on Random Forest (RF-RFE) is applied and has eliminated five variables. By machine learning techniques namely SVR, the best model is SVR with kernel Radial Basis Function (RBF) with complete feature. The best model is used to predict compressive strength in April 2019 which has the best RMSE criteria of 8.78.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSSt 519.536 Feb p-1 2019 |
| Uncontrolled Keywords: | Kuat Tekan, Multikolinearitas, PCC, PCR, RBF, RFE, RMSE, SVR. |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression Q Science > QA Mathematics > QA353.K47 Kernel functions (analysis) Q Science > QC Physics > QC173.4.C63 Composite materials |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Rizki Febriasto |
| Date Deposited: | 27 Dec 2022 01:49 |
| Last Modified: | 27 Dec 2022 01:49 |
| URI: | http://repository.its.ac.id/id/eprint/63402 |
Actions (login required)
 |
View Item |