Please use this identifier to cite or link to this item:
|j48 Classifier for Software Effort Estimation
|Software Engineering, Decision Tree Based Approaches, j48 classifier, Random Forest Classifier
|The software effort estimation is the technique, which is applied to estimate efforts for the software development. The software effort estimation is the challenging task due to large size of the software. The models, which are designed for the effort estimation depend upon the KLOC (Kilo Lines of Code) values. The incorrect estimation of the KLOC value leads to incorrect estimation of the effort, which raises the MRE (Mean Relative Error) value. The COCOMO (Constructive Cost Model) model is the most popular model for software’s effort estimation. The effort calculated in COCOMO model is directly proportional to KLOC value. This work is based on the KLOC value estimation for the reduction of MRE. In the existing technique, random forest classifier is applied for the reduction of MRE value. In this thesis work, j48 model is applied for the reduction of MRE value. The proposed and existing techniques are implemented in Python using Anaconda software. The performance of both algorithms is compared in terms of accuracy and execution time. The j48 algorithm shows high accuracy and less execution time for the effort estimation.
|Appears in Collections:
Files in This Item:
|801631014_Nancy Sharma_ME Thesis.pdf
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.