| DC 欄位 |
值 |
語言 |
| DC.contributor | 資訊管理學系在職專班 | zh_TW |
| DC.creator | 傅昱瑋 | zh_TW |
| DC.creator | Yu-Wei Fu | en_US |
| dc.date.accessioned | 2021-8-10T07:39:07Z | |
| dc.date.available | 2021-8-10T07:39:07Z | |
| dc.date.issued | 2021 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=108453030 | |
| dc.contributor.department | 資訊管理學系在職專班 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 本研究之主要目的是建置求職者在職時間預測模型，運用離職員工的人事資料和人格測驗去探討和分析出離職天數，未來將可提供主管們在面試時，在評比求職者的標準中可將此研究結果納入參考數據的一種，進而可篩選出不易離職的員工，以降低員工離職率和公司成本的相關影響，且可提升公司的信譽和競爭力。
本研究所採用的研究方法是選擇類神經網路(Neural Network)演算法建置適用於的在職天數預測模型(PCWD Model(Predict Candidates Work Days Model)，設計四個實驗方法，首先最普遍常見的是運用離職員工的人事資料是否會影響在職天數的長短，第二是離職員工的人格測驗結果是否會影響在職天數的長短，第三是離職員工的人事資料和人格測驗兩個資料集整併是否會影響在職天數的長短，最後是將人事資料的預測結果和人格測驗的預測結果，將兩者預測結果進行預測研究是否會影響在職天數的長短，依實驗方法進行假說和擬定了設計出四個預測模型來驗證。
本研究設計出四個預測模型之實驗預測結果依序上述依此類說明預測結果，人事資料預測模型，預測結果的準確率為0.88，人格測驗預測模型，預測結果的準確率為0.99，人事資料和人格測驗預測模型，預測結果的準確率為0.91，人事資料結果和人格測驗結果預測模型，預測結果的準確率為0.89，由實驗結果的數據可得知本研究的PCWD預測模型能有效地降低員工離職率和公司成本的問題。 | zh_TW |
| dc.description.abstract | The main purpose of this research is to build a predictive model of job applicants′ on-the-job time, and use the personnel data and personality tests of the employees to explore and analyze the number of days left. Incorporating the results of this research into a type of reference data can then screen out employees who are not easy to quit, so as to reduce the impact of employee turnover and company costs, and improve the company′s reputation and competitiveness.
The research method adopted in this research is to select a Neural Network algorithm to build a suitable PCWD Model (Predict Candidates Work Days Model). Four experimental methods are designed. First of all, the most common It is whether the use of personnel data of the resigned employee will affect the length of the number of days on the job, the second is whether the result of the personality test of the resigned employee will affect the length of the number of days on the job, and the third is whether the two data sets of the personnel data and the personality test of the resigned employee are integrated It will affect the length of working days. Finally, the prediction results of the personnel data and the prediction results of the personality test are used to predict whether the two prediction results will affect the length of the working days. The hypotheses are based on experimental methods and four designs are drawn up, and predictive model to verify.
In this study, the experimental prediction results of the four prediction models were designed in the order described above and so on to explain the prediction results. The personnel data prediction model has an accuracy rate of 0.88. The personality test prediction model has an accuracy rate of 0.99. Personnel data and the personality test prediction model, the accuracy rate of the prediction result is 0.91, and the prediction model of personnel data and personality test results has an accuracy rate of 0.89. The data of the experimental results shows that the PCWD prediction model of this study can effectively reduce Employee turnover rate and company cost issues. | en_US |
| DC.subject | 員工離職 | zh_TW |
| DC.subject | 人格特質 | zh_TW |
| DC.subject | 類神經網路 | zh_TW |
| DC.subject | Employee turnover | en_US |
| DC.subject | Character | en_US |
| DC.subject | Neural network | en_US |
| DC.title | 應用類神經網路於求職者在職時間預測模型之研究 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Predict Candidates Work Days Model Based on Neural Network | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |