應用資料採礦技術於大學畢業生轉職傾向之預測研究

根據主計總處105年人力運用調查報告中可看出年輕族群為高轉職傾向的一個群體。本研究參考過往對於大學新鮮人轉換工作的特質與傾向等相關文獻，歸納出導致轉換工作的原因包含薪資、學歷、專業知識、專業證照與技能、學以致用程度等；並應用教育部之畢業生流向調查的資料，透過國內某大學於2016年針對該校101學年度畢業生在畢業三年後之流向調查資料，以統計軟體SAS-EM（Enterprise Miner）中決策樹（Decision Tree）、類神經網路（Artificial Neural Network）、支援向量機（Support Vector Machine）等資料採礦技術進行分析，期望能建立合適的統計模型，進行畢業生轉職傾向之預測，再與此批101學年度畢業生於2014年進行的畢業後一年之畢業生資料相互比對，探討這些學生經過兩年時間後，其工作場域是否有發生轉變，最後，利用上述資料進行數據建模，並評估本研究各種預測模式之準確率。

關鍵字

轉職傾向；資料採礦技術；決策樹；類神經網路；支援向量機

並列摘要

According to the 105-year manpower utilization survey, young generation are identified as highly motivated on job shifting. Based on the past researches, this study summarizes the reasons lead to the job shifting including salary, education level, professional knowledge, certification and license. According to the survey from a domestic university in 2016, this is for the students graduate from the university for three years, this study will analyze the data by using Decision Tree, Artificial Neural Network, and Support Vector Machine in SAS-EM. Based on the students' data and the past research, this study hopes to build a model to predict the tendency of job shifting on the graduation student. After the prediction, comparing with the same student's survey in 2014 to find out if their job has changed or not, then assess the model's accuracy.

並列關鍵字

Job Shifting Tendency ； Data Mining Technology ； Decision Tree ； Artificial Neural Network ； Support Vector Machine

主題瀏覽