INTRODUCTION Dataset The Artificial COCOMO 81 dataset Published
INTRODUCTION (Dataset) ﺩیﺘﺎﺳﺖ آﻮﺭی ﺟﻤﻊ ﻣﻌﺘﺒﺮ The Artificial COCOMO’ 81 dataset • • Published By Boehm In His Seminal Book "Software Engineering Economics" COCOMO’ 81 Dataset Containing 63 Software Projects including business, scientific and system projects, described by 19 variables (18 input variables and 1 output variable) © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION Variables of the COCOMO data set © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION The Tukutuku dataset • • contains 53 Web projects Each Web application described using 9 numerical attributes such as: the number of html or shtml files used, the number of media files and team experience © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION Software attributes for Tukutuku dataset © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION (Dataset) ﺩیﺘﺎﺳﺖ آﻮﺭی ﺟﻤﻊ ﻣﻌﺘﺒﺮ The Desharnais dataset • • The Desharnais data set consists of 81 software projects described by 11 variables nine independent and two dependent © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION ﻣﻨﺎﺑﻊ ﺳﺎیﺮ NASA Albrecht Kemerer © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
WHY SBSE ﻧﻤﺎیﺶ ﻫﺎی ﺭﻭﺵ ﺳﺎیﺮ Tree String Bit-matrix Integer array Integer vector Binary String Object-based Matrix Vector © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION کﺎﺭ ﺍﺭﺯیﺎﺑی (Software ) ﺑﺮﺍی ﺍﺭﺯیﺎﺑی ﻣﺪﻝ ﻫﺎی ﺗﺨﻤیﻦ ﺗﻼﺵ ﻧﺮﻡ ﺍﻓﺰﺍﺭ ، ﺑﻪ ﻋﻨﻮﺍﻥ ﻣﺜﺎﻝ effort estimation models • Magnitude Of Relative Error (MRE) • The MRE values are calculated for each project in the datasets © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION • Mean Magnitude Of Relative Error (MMRE) MMRE computes the average over N projects • • Generally, the acceptable target value for MMRE is 25% his indicates that on the average, the accuracy of the established estimation models would be less than 25%. Another widely used criterion is the Pred(l) which represents the percentage of MRE that is less than or equal to the value l among all projects © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION • The definition of Pred(l) is given as follows: Where N is the total number of observations and k is the number of observations whose MRE is less or equal to l • A common value for l is 0. 25, which also used in the present study • The Pred(0. 25) represents the percentage of projects whose MRE is less or equal to 25% • The Pred(0. 25) value identifies the effort estimates that are generally accurate whereas the MMRE is fairly conservative with a bias against overestimates. • © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION EXP 01 GA-based method for feature selection and parameters optimization for machine learning regression applied to software effort estimation © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
INTRODUCTION کﺎﺭ ﺍﻧﺠﺎﻡ ﻣﺮﺍﺣﻞ (Feature Extraction) ﻫﺎ ﻭیژگی ﺍﺳﺘﺨﺮﺍﺝ (Feature Selection) ﻫﺎ ﻭیژگی ﺍﻧﺘﺨﺎﺏ (Classification) ﺑﻨﺪی ﺩﺳﺘﻪ © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
(. . . )ﺍﺩﺍﻣﻪ ﺍﺭﺯیﺎﺑی : ﺗﻌﺪﺍﺩ ﻭیژگی ﻫﺎی ﻭﺭﻭﺩی ﺍﻧﺘﺨﺎﺏ ﺗﻮﺳﻂ ﺍﻟگﻮﺭیﺘﻢ ﻫﺎی ﺩﺳﺘﻪ ﺑﻨﺪ ﻣﻮﺭﺩ ﻣﻄﺎﻟﻌﻪ • • SVR RBF – 4. 4 input features on average. SVR linear – 3. 3 input features on average. MLP – 2. 7 input features on average. M 5 P – 4. 9 input features on average. © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
(. . . )ﺍﺩﺍﻣﻪ • • • ﺍﺭﺯیﺎﺑی Each attribute of the project is categorized according with level of impact on effort estimation The levels are: Very Low, Nominal, High, Very High and Extra High ach one of these levels has a numerical value associated After preprocessing the data set, we divided it into six different pairs of training sets and test sets © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
(. . . )ﺍﺩﺍﻣﻪ ﺍﺭﺯیﺎﺑی COCOMO data set: pairs of training sets and test sets. Next, we used these six pairs of training and test sets for training and testing the proposed method and to compare the results obtained with those of Tronto et al. © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
(. . . )ﺍﺩﺍﻣﻪ ﺍﺭﺯیﺎﺑی Average and standard deviation of experimental results for COCOMO data set using GA-based approach © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
(. . . )ﺍﺩﺍﻣﻪ ﺍﺭﺯیﺎﺑی Average and standard deviation of experimental results for COCOMO data set using GA-based approach © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
(. . . )ﺍﺩﺍﻣﻪ ﺍﺭﺯیﺎﺑی Comparison of the GA-based method with MLP [26] on the COCOMO data set. © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
Thanks © Copyright Showeet. com This work is licensed under a Creative Commons Attribution
- Slides: 52