Presentation 2012-01-26
Evaluation of Identification of Variables to Transform into Logarithmic in Cost Prediction
Kenta MIYOSHI, Katsuma UBUKATA, Takeshi KAKIMOTO, Shinji KUSUMOTO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) With the development of large and complex software, it may become difficult to proceed the project according to designed, due to its "quality", "cost", and "delivery". In software development, effort has been made to predict the schedule of administration. The prediction accuracy is improved by logarithmic transformation of variables, which is used to predict. However, it is not effective to transform all variables into logarithmic. It takes too much time and not realistic to check for each variable whether it should be transformed or not. It is efficient to select important variables which are expected to improve the prediction accuracy by transform into logarithmic compared to the case we transform for comparison. In the proposed method, we divided whole project into small sections, and determined the inclination in each divided segment, and decide whether the variable should be transform into logarithmic or not by the continuousness of increasing inclination. As a result, better prediction accuracy was obtained by our method, and the accuracy was higher compared to the case of techniques to compare.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Effort Prediction / Multiple regression analysis / Normal distribution / Kurtosis / Skewness
Paper # MSS2011-56,SS2011-41
Date of Issue

Conference Information
Committee SS
Conference Date 2012/1/19(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Software Science (SS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Evaluation of Identification of Variables to Transform into Logarithmic in Cost Prediction
Sub Title (in English)
Keyword(1) Effort Prediction
Keyword(2) Multiple regression analysis
Keyword(3) Normal distribution
Keyword(4) Kurtosis
Keyword(5) Skewness
1st Author's Name Kenta MIYOSHI
1st Author's Affiliation Department of Electrical and Computer Engineering, Kagawa National College of Technology()
2nd Author's Name Katsuma UBUKATA
2nd Author's Affiliation Nomura Research Institute
3rd Author's Name Takeshi KAKIMOTO
3rd Author's Affiliation Department of Electrical and Computer Engineering, Kagawa National College of Technology
4th Author's Name Shinji KUSUMOTO
4th Author's Affiliation Graduate School of Information Science and Technology, Osaka University
Date 2012-01-26
Paper # MSS2011-56,SS2011-41
Volume (vol) vol.111
Number (no) 406
Page pp.pp.-
#Pages 6
Date of Issue