Course Scheduling for Employee Training Using Data Mining

• Main Authors: ThiTamNguyen 阮氏心, ThiTamNguyen
• Other Authors: Chien-Wen Wu 吴建文
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/sfbr6d

碩士 === 國立臺北科技大學 === 管理國際學生碩士專班 (IMBA) === 100 === In this research, we study the problem of scheduling a training course for an enterprise assuming that employees are busy and may not be available at all the time. We want to schedule several time slots for the course so that employee constraints, the instructor constraints and the room constraints can all be satisfied. A mathematical model is provided for the problem. Also, an algorithm based on Frequent Itemset Mining (FIM) is presented for the problem. The experiments were performed on a 1.2 GHz PC with 2 GB of memory running Windows XP. For our approach, we employed a version of the Mafia algorithm for Frequent Itemset Mining. For the exact approach, we used Visual C++ and the CPLEX callable library to solve our mathematical model for the comparison purpose. As a result, our approach performs faster than the CPLEX approach in overall. The maximum improved ratio is 2589.88%. However, when CF equals to 0.7, our approach shows poorer results. It may be explained by the influence of the availability of employees on these two approaches.