Software cost estimation based on modified <Emphasis Type="Italic">K</Emphasis>-Modes clustering Algorithm |
| |
Authors: | Partha Sarathi Bishnu Vandana Bhattacherjee |
| |
Affiliation: | 1.Birla Institute of Technology,Ranchi,India |
| |
Abstract: | Unsupervised technique like clustering may be used for software cost estimation in situations where parametric models are difficult to develop. This paper presents a software cost estimation model based on a modified K-Modes clustering algorithm. The aims of this paper are: first, the modified K-Modes clustering which is an enhancement over the simple K-Modes algorithm using a proper dissimilarity measure for mixed data types, is presented and second, the proposed K-Modes algorithm is applied for software cost estimation. We have compared our modified K-Modes algorithm with existing algorithms on different software cost estimation datasets, and results showed the effectiveness of our proposed algorithm. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|