برنامه ریزی دوره های تعلیم آموزش های داده کاوی: به عنوان نمونه استفاده از شرکت موتور چینی

In Taiwan, most industries are of small and medium scale, and there are limited resources for educational training. Increasing the quality of personnel by cultivating talents for the future becomes an extremely important issue. With the growth of firms and the increase in their needs, the database is also growing. We should therefore determine how to recognize and extract the useful information contained in this database in order to apply it in such a way that assists companies in meeting their increasing and changing needs. This research collects data of personnel educational training in China Motor Corporation by cluster analysis, decision tree algorithm and back-propagation neural networks for mining analysis and classification. Based on the algorithm classification result, we finally propose the demand model suitable for educational training in other related industries. The research is expected to explore how to maximize results through planning the courses and the personnel’s participation in the training. We try to determine the key factors essential to the success of educational training. Once identified, this information can then serve as the basis for other firms’ future planning of educational training strategies with regard to innovation and breakthrough.

Georgenson (1982) noted that although the funds of corporate educational training continued growing, only 10% of the training was effective. For innovation and breakthrough, this paper conducted a random sampling of 38,000 pieces of data in the personnel educational training database of China Motor Corporation according to employees’ positions and departments. This was done through the decision tree algorithm developed upon the principle of Data Mining for study, mining and analysis in order to find the relation between the classification of educational training courses and the classification of the employees’ occupation, the courses favored by the employees of different occupations and their accomplishment probability. However, in the educational training, the employees’ characteristics or positions will influence the time spent on providing employee education and the funds invested in education. Thus, the educational training should be precise. Since educational training tends to be conducted by the human resources department, the cognitive difference between the employees’ characteristics and the human resources department might lead to insufficient funds or the waste of the investment. It will take more time to adjust the administration’s strategies when there is an imbalance of manpower supply and demand. Thus, manpower planning and the precision of prediction will influence corporate efficiency and quality. Excess manpower supply will lead to idleness and inefficient use of personnel and indirectly result in the increase of the cost and the waste of resources; on the contrary, insufficient manpower will lead to imbalanced distribution of resources and reduce the corporate quality. Thus, with complicated training courses and plenty of employees, we should determine how to classify educational training precisely and meet the manpower demand to increase educational training quality and reduce excess expenses. This research expects to effectively analyze and evaluate by database in order to modify the original educational training model, and to construct an educational training system through manpower forecasting. It will also construct an evaluation model to formulate the most accurate forecasting models as the basis for designing employee educational training and to increase the precision of employee classification and reduce the waste of human resources. More than 95% of the industries in Taiwan are of small and medium scale, and the firms have limited resources for educational training (Farh, 1995). Nonetheless, the funds for corporate educational training continue growing, and in order to improve human resources quality, we should cultivate employees’ talents for the future. This research follows the “Input-Process-Output” (IPO) proposed by Bushnell (1990) as the basic framework of the educational training model, as shown in Fig. 1.The research purposes are reorganized below: (1) We construct a “training input, practice training process and training effect” assessment model to explore an overview of the training effect. (2) We explore the possible mediating effect between training input and effect during the practice. (3) We validate the feasibility of “training input, practice training process and training effect” assessment model in small- and medium-sized enterprises within the manufacturing industry in Taiwan. (4) The educational training fund invested in the firm is one of the critical indexes for assessing “training input”. We try to find the key indexes of training funds invested in the firms to explore the relationship among training funds, the training process and the training effect. (5) We plan educational training strategies specially adapted to meet the specific needs of employees in different positions.

Based on the results of cluster analysis and decision tree algorithm, this paper draws the following conclusions. First of all, according to cluster analysis, we find that there are more production and R&D courses in the China Motor Corporation and fewer marketing, human resources and finance courses. We suggest that the company increase the number of courses in the latter category. Employees in all departments can participate in the planning of general courses, and the advanced study and core courses can be much better organized. Employees may not be satisfied with their present work and therefore intend to learn other skills to change their occupation. Provided they are aware of this situation, supervisors can focus on meeting the career-building needs of their workers, with the goal of retaining their employees. With cluster analysis, we can indicate the similarity of certain courses with the goal of arranging a series of courses in order to provide effective educational training, encourage the employees to systematically join in the courses, increase the qualification rate, match the work content with the employees’ profession and aspirations and increase work efficiency. Finally, according to decision tree algorithm, we find that human resources, finance, production and R&D courses in an educational training center are advantageous. The participants in the educational training are the high-ranking supervisors. Based on the above two kinds of algorithms, we find that different algorithms show the information of different constructs. The decision makers can thus control the current situation of employee performance through continuing education. It also shows the precision of the attributes among the variables in this paper. We apply data mining and try to find the correlation between the course classification of educational training and the positions of those employees participating in educational training from the database of the China Motor Training Center. We first recognize the cluster relation among the above three variables by cluster analysis. According to the Tree Diagram, we identify the similarity of certain courses in the China Motor Training Center in order to develop a series of training courses; we further find the decision tree diagram of three variables by CHAID in the decision tree algorithm. According to the tree diagram, we find the most favored, prestigious courses in the China Motor Training Center’s training and the preferred training courses of the employees in different positions. We also realize that the qualification rate of those joining in training courses of different classifications is higher. We indicate the correlation between the classification of training courses and the employees’ positions, and between the favorite course classification of the employees with different positions and their accomplishment rate. Based on the results of the two kinds of algorithms, China Motor Training Center can plan its training courses according to its employees’ different positions and can construct prestigious, effective educational training courses with high efficiency.