Optimization of analogy weights by genetic algorithm for software effort estimation

This research work is aimed at to provide effective cost estimation methodology emphasize on cost effort and time . This paper summarizes the cost effort estimation of most conventionally used models like organic and semi-detached models using an improved version of genetic algorithm that enhances an empirical methodology to reduce the cost factor and time factor in software projects. Constructive cost model(Cocomo model) is broadly used for the fruitful valuation of cost estimation which is based on KLOC method(thousands of lines of code).This method yields beneficial result in case of lines of code method but lacks in terms of concept and logics. The same is estimated directly and is computed using the function point analysis. In the software development lifecycle, the software cost effort estimation is the most demanding process. The accuracy of the estimate in choosing the estimation model is an essential factor. Such conventional software effort estimation techniques fail to compute the accuracy of effort estimation and it is not up to the mark. So here, we tend to propose the cost reduction in the software projects by using the improved version of the known genetic algorithm.

Download Full-text

Genetic Algorithm and Support Vector Regression for Software Effort Estimation

Advanced Materials Research ◽

10.4028/www.scientific.net/amr.282-283.748 ◽

2011 ◽

Vol 282-283 ◽

pp. 748-752 ◽

Cited By ~ 1

Author(s):

Jin Cherng Lin ◽

Chu Ting Chang

Keyword(s):

Genetic Algorithm ◽

Support Vector ◽

Effort Estimation ◽

Software Effort Estimation ◽

Software Developers ◽

Factors Affecting ◽

Proposed Model ◽

Vector Machines ◽

Best Parameter ◽

The Cost

For software developers, accurately forecasting software effort is very important. In the field of software engineering, it is also a very challenging topic. Miscalculated software effort in the early phase might cause a serious consequence. It not only effects the schedule, but also increases the cost price. It might cause a huge deficit. Because all of the different software development team has it is own way to calculate the software effort, the factors affecting project development are also varies. In order to solve these problems, this paper proposes a model which combines genetic algorithm (GA) with support vector machines (SVM). We can find the best parameter of SVM regression by the proposed model, and make more accurate prediction. During the research, we test and verify our model by using the historical data in COCOMO. We will show the results by prediction level (PRED) and mean magnitude of relative error (MMRE).

Download Full-text

Estimating Software Development Efforts Using a Random Forest-Based Stacked Ensemble Approach

Electronics ◽

10.3390/electronics10101195 ◽

2021 ◽

Vol 10 (10) ◽

pp. 1195

Author(s):

Priya Varshini A G ◽

Anitha Kumari K ◽

Vijayakumar Varadarajan

Keyword(s):

Deep Learning ◽

Random Forest ◽

Software Development ◽

Weighted Averaging ◽

Software Project ◽

Effort Estimation ◽

Software Effort Estimation ◽

Single Model ◽

Ensemble Techniques ◽

Project Estimation

Software Project Estimation is a challenging and important activity in developing software projects. Software Project Estimation includes Software Time Estimation, Software Resource Estimation, Software Cost Estimation, and Software Effort Estimation. Software Effort Estimation focuses on predicting the number of hours of work (effort in terms of person-hours or person-months) required to develop or maintain a software application. It is difficult to forecast effort during the initial stages of software development. Various machine learning and deep learning models have been developed to predict the effort estimation. In this paper, single model approaches and ensemble approaches were considered for estimation. Ensemble techniques are the combination of several single models. Ensemble techniques considered for estimation were averaging, weighted averaging, bagging, boosting, and stacking. Various stacking models considered and evaluated were stacking using a generalized linear model, stacking using decision tree, stacking using a support vector machine, and stacking using random forest. Datasets considered for estimation were Albrecht, China, Desharnais, Kemerer, Kitchenham, Maxwell, and Cocomo81. Evaluation measures used were mean absolute error, root mean squared error, and R-squared. The results proved that the proposed stacking using random forest provides the best results compared with single model approaches using the machine or deep learning algorithms and other ensemble techniques.

Download Full-text