Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476500Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.author | Sophia Jamila Zahra | - |
| dc.date.accessioned | 2023-10-06T09:19:41Z | - |
| dc.date.available | 2023-10-06T09:19:41Z | - |
| dc.date.issued | 2011 | - |
| dc.identifier.other | ukmvital:114684 | - |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476500 | - |
| dc.description | Now a days applying just in time business is demanding in a manufacturing company. The key issue in practicing just in time business is always having an optimum stock balance that align with the production planning. Therefore, accurate demand forecast is necessary to ensure existing resources being used as efficient as possible to accomplish a production plan. Conventionally, statistical approaches such as exponential smoothing techniques are used in demand forecasting problems. However, this study aims to improve the accuracy of forecast demand in production planning by using artificial intelligence techniques, such as neural network approach and genetic programming. A real industrial data have been used in the experiments. It is from a company that supplying various engineering plastic components especially for automotive door latches. The data consists of weekly real product quantity and generated forecast values by using exponential smoothing. In exponential smoothing, alpha value is a parameter determining the performance of the forecast. In this study, the Neural Network and Genetic Programming are used to evaluate the performance of different alpha value. There are five different alpha values, 0.1 to 0.5. There are two parts of experiment: 1) A combination of alpha values and 2) A separated alpha value. In the first part, generated forecast values for each alpha are combined as an input in training and testing network. Meanwhile, in the second part Neural Network evaluated the separate alpha value as an input and Genetic Programming model combined alpha values as an input. Variance and MSEP are chosen to measure the performance for these two techniques. In variance testing phase, the best result of Neural Network improved 98.70% and Genetic Programming improved 92.56%. And in MSEP results, the Neural Network has reached the much smaller error with 2.35E+04 and Genetic Programming succeed achieved 7.98E+04 error. Following by these results indicate that Neural Network was able to obtained best performance, near forecast rather than Genetic Programming.,“Certification of Master’s/Doctoral Thesis” is not available,Master Information Technology | - |
| dc.language.iso | eng | - |
| dc.publisher | UKM, Bangi | - |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | - |
| dc.rights | UKM | - |
| dc.subject | Futures market -- Forecasting -- Mathematical models | - |
| dc.subject | Automatic control -- Data processing | - |
| dc.subject | Neural networks (Computer science) | - |
| dc.subject | Mathematical statistics -- Data processing | - |
| dc.subject | Soft computing | - |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | - |
| dc.subject | Dissertations, Academic -- Malaysia | - |
| dc.title | Production demand forecasting model using neural network and genetic programming | - |
| dc.type | theses | - |
| dc.format.pages | 105 | - |
| dc.identifier.callno | QA76.87.S668 2011 3 tesis | - |
| dc.identifier.barcode | 002391(2011) | - |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_114684+SOURCE1+SOURCE1.0.PDF Restricted Access | 2.38 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.