Ramadhan, Defa Ihsan (2020) Analisis Kinerja Peramalan Dan Klasifikasi Permintaan Part Otomotif Dengan Pendekatan Time-Series Dan Data Mining. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
02411640000032-Undergraduate_Thesis.pdf Download (5MB) | Preview |
Abstract
Kontribusi after sales service yang besar terhadap profit dan pertumbuhan bisnis menyebabkan spare part management menjadi faktor yang penting untuk bersaing di pasar. Termasuk bagi PT. X, produsen mobil penumpang dengan market share terbesar secara global di tahun 2019. Berbagai upaya spare part management telah dilakukan PT. X untuk memenuhi kebutuhan part domestik maupun ekspor. Salah satu upayanya adalah melakukan perencanaan produksi bulanan (Getsudo), termasuk peramalan permintaan spare part setiap bulannya. Akan tetapi, metode peramalan moving average, yang saat ini digunakan untuk semua spare part di PT. X, kurang efektif pada beberapa pola permintaan spare part yang variasinya tinggi. Penyimpangan pada hasil peramalan berdampak pada lead time back order dan biaya material handling yang semakin besar. Berdasarkan hal tersebut maka PT. X memerlukan perbaikan terhadap sistem peramalan spare part-nya. Tugas akhir ini memiliki dua tujuan utama. Tujuan pertama adalah mengusulkan pengelompokkan spare part berdasarkan pola permintaannya sebelum diramalkan. Tujuan kedua adalah menentukan metode peramalan yang paling sesuai untuk masing-masing kelompok spare part dengan cara membandingkan empat metode peramalan yaitu: Croston, Modifikasi Croston, SVR, dan ANN. Seluruh metode peramalan dibandingkan berdasarkan parameter forecasting error dan robustness. Hasil penelitian ini menunjukkan bahwa metode SVR dan ANN memiliki kinerja yang lebih unggul dari metode lainnya di tahap training maupun testing. Selain itu, ketika diimplementasikan untuk peramalan multi-periode, metode SVR dan ANN juga lebih unggul dan dapat memperbaiki kesalahan peramalan sebesar 18.8% dibandingkan dengan metode peramalan yang saat ini digunakan oleh PT. X.
====================================================================================================
The large contribution to profit and business growth of after sales service have made spare part management as an important factor to compete in the market. Especially in PT. X, a passenger car manufacturer with the largest market share globally in 2019, various part of service management efforts have been carried out to meet the domestic and export part demand. One of them is monthly production planning (Getsudo), in which including forecasting spare part demand every month. However, the moving average forecasting method, which is currently being used to forecast all spare part demand in PT. X, is proven less effective to high demand variations of spare part. Deviations in forecasting results in greater lead time backorders and material handling costs. Because of those problems, PT. X requires improvements to its spare part forecasting system. This final project has two main objectives. The first objective is to propose the classification system of spare part based on the demand pattern before being forecasted. The second objective is to determine the forecasting method that is most appropriate for each group of spare part by comparing the four forecasting methods, namely: Croston, Croston Modification, SVR, & ANN. All forecasting methods are compared based on forecasting error & robustness. The results of this study indicate that the SVR and ANN methods have performance that is superior to other methods in the training and testing data. In addition, when implemented for multi-period forecasting, the SVR and ANN methods are also superior and can improve forecasting errors by 18.8% compared to the existing forecasting method.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Artificial Neural Network, Croston’s Method, Data Mining, Spare Part Forecasting, Support Vector Regression, Time-Series Forecasting, Artificial Neural Network, Data Mining, Metode Croston, Peramalan Spare Part, Peramalan Time-Series, Support Vector Regression. |
| Subjects: | Q Science > QA Mathematics > QA278.55 Cluster analysis Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science) |
| Divisions: | Faculty of Industrial Technology > Industrial Engineering > 26201-(S1) Undergraduate Thesis |
| Depositing User: | Defa Ihsan Ramadhan |
| Date Deposited: | 14 Aug 2020 00:20 |
| Last Modified: | 09 Jun 2023 14:58 |
| URI: | http://repository.its.ac.id/id/eprint/78047 |
Actions (login required)
 |
View Item |