姓名 |
盧彥夫(Yeng-fu Lu) |
電子郵件信箱 |
s9111638@mail.cyut.edu.tw |
畢業系所 |
營建工程系碩士班(Department
and Graduate Institute of Constrction Engineering) |
畢業學位 |
碩士(Master) |
畢業時期 |
92學年第2學期 |
論文名稱(中) |
類神經網路與學習曲線預測學習效應成效比較之研究 |
論文名稱(英) |
Comparisons of
Prediction Efficiency of Learning Effect between Neural Network and
Learning Curve |
檔案 |
本電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。 請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。
|
論文使用權限 |
校內外都兩年後公開 |
論文語文/頁數 |
中文/62 |
摘要(中) |
在營建工程的領域中,時常可見應用學習曲線做重複性工程的數據分析或是工期預測,由於應用學習曲線的方法,從過去的文獻中得知,在前期的資料預測上能達到較佳的預測效果,但對後期資料而言,則容易產生較大的預測誤差。本研究利用類神經網路的倒傳遞模式,以調整輸入值與輸出值之間的權重值及閥值,使輸入值與輸出值產生相對應的關係,以求能達成預測的效能,並以預鑄節塊懸臂工法橋樑架設工程來做實際案例的驗證,研究發現類神經網路確實能預測重複性工程上的工期預測。 |
摘要(英) |
In a repeatable
construction project, same activity is repeatedly performed. Each time
the activity is performed, the workers presumably discover how to make
it better and quicker and such phenomena is called learning effect. The
learning effect is usually forecasted by learning curves for the sake of
predicting completion time for a project. However, learning curves are
usually more accurate for forecast duration in early construction stage
than later one of a project. This research aims at comparing the
efficiency of applying neural network and learning curves for the
prediction of learning effect. Case study shows neural network providing
a better quality in the forecasting learning phenomena. |
關鍵字(中) |
學習曲線
類神經網路 |
關鍵字(英) |
learning curve
neural network |
指導教授 |
鄭道明
|