PENERAPAN METODE NAIVE BAYES CLASSIFICATION UNTUK PENENTUAN KELAS ON TIME KESESUAIAN WAKTU JOB ORDER PERBAIKAN KOMPONEN ALAT BERAT (Studi Kasus: PT. Komatsu Remanufacturing Asia Cabang Balikpapan)
| Main Authors: | , Syamsul Bahri, , Dr. Azhari SN, Drs., MT. |
|---|---|
| Format: | Thesis NonPeerReviewed |
| Terbitan: |
[Yogyakarta] : Universitas Gadjah Mada
, 2014
|
| Subjects: | |
| Online Access: |
https://repository.ugm.ac.id/127695/ http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=67965 |
Daftar Isi:
- Every manufacturing companies that moves in the job order system with a wide network realizes the importance of customer service that meets the requirements to win the market competition, scheduling is an important element in the fulfillment of orders (order) on time. However, swelling of the amount of data a company has a job order is not followed by an increase in the ability of the company to classify the likelihood that completion of job order can be handled on time or not. Therefore the solution is to do with data mining techniques. Data mining techniques used in this study for classification timeliness of completion of job order repairs heavy equipment components by applying a Naive Bayes Classification algorithm. This algorithm is a classification algorithm in data mining refers to the Bayes theorem is simple but has a high accuracy. Preprocessed initial-job order data, processed using Naive Bayes algorithm to construct the model, then the model is used to classify the data into a new job order class on time or not on time. Testing techniques the accuracy of the model was measured by 10-fold cross validation, and shows that the smallest value of accuracy is 88.12% produced in the 5000 data sample, and produces the largest value of accuracy 95.80% on a data sample of 1000. The test results with Rapid Miner 5.3 software using Naive Bayes method obtained the smallest values of accuracy 97.17% with a sample of 30000 and 100% the value of the highest accuracy with a sample of 200. And for the method of SVM obtained the smallest value is 92.20% accuracy with sample 500 and the highest accuracy is 99.22% with a sample of 5000.