查詢結果分析
來源資料
頁籤選單縮合
題名 | 利用類神經-模糊理論評定契合程度--以管理人員甄選為例=An Alternative Approach to Assess Fit by Using Neuro-Fuzzy--An Example of Managerial Personnel Selection |
---|---|
作者 | 林金賢; 許碧芬; 鄭妃君; | 書刊名 | 管理學報 |
卷期 | 19:1 2002.03[民91.03] |
頁次 | 頁77-108 |
分類號 | 494.311 |
關鍵詞 | 遴選; 模糊邏輯; 類神經網路; 管理知能模式; 類神經模糊; Selection; Fuzzy logic; Neural network; Competency model; Neuro-fuzzy; |
語文 | 中文(Chinese) |
中文摘要 | 傳統的人力資源管理強調以工作分析為基礎,進而發展人力資源的各個功能 (human resource practice)。而現代的人力資源管理則強調以知能 (competency) 為核心建構策略性人力資源管理,強調組織策略與內部整體人力資源功能間的外部契合 (external fit) 與組織內部各個人力資源功能彼此間的內部契合 (internal fit)。然而在過去探討契合的文獻中,從 selection approach到 interaction approach,一直到 system approach,由於各個方法的前提假設不同,因而其驗證的方式與結論也有所差異。本研究希望從契合的觀念出發,以策略性人力資源管理中的人員甄選為例,利用類神經網路與模糊理論對非線性關係的捕捉能力來配置有關契合程度的模糊專家系統,以突破過去方法上的限制與不足,提供管理者一有效的人員甄選決策模式。除了用對績效的預測能力來取代過去統計方法上對線性關係檢定的顯著性,給予契合另一層面的詮釋外,此專家系統亦可用來對企業內現有的人員進行知能診斷,提供未來人力發展,教育訓練的方向,以及績效評估的參考指標。另外,由於此專家系統可以隨著時間的變動自我學習、自我調整,因此即使構面間的契合關係隨著時間在變動,此模式仍然可以適用。 |
英文摘要 | Traditionally the human resource practices are developed based on job analysis. Nowadays the strategic human resource management, based on competency model, are emphasizing on external fit and internal fit. However, from the studies about fit, there are still a lot of unsolved problems among the existing approaches, selection approach, interaction approach, and system approach, due to their assumptions and methods. This research, based on the idea of fit, is trying to construct an expert system to select the promising manager by using neuro-fuzzy technique. It is trying to breakthrough the constraints and limits of the traditional statistical technique by using the validity of the forecasting results of the expert system to replace the statistical significance of the variables, giving an alternative approach to explain fit. This expert system can also be used as the competency diagnosis for people in the organization, providing clues to personnel development and career training. Besides this system can still be used even if the relationship among these factors are changing as time changes due to the self-learning capability of this system。 |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。