Ensemble Learning Approach for Effective Software Development Effort Estimation with Future Ranking

To provide a client with a high-quality product, software development requires a significant amount of time and effort. Accurate estimates and on-time delivery are requirements for the software industry. The proper effort, resources, time, and schedule needed to complete a software project on a tigh...

Descripción completa

Detalles Bibliográficos
Autores: Eswara Rao, K., Pydi, Balamurali, Annan Naidu, P., Prasann, U. D., Anjaneyulu, P.
Tipo de recurso: artículo
Fecha de publicación:2023
País:España
Institución:Universidad de Salamanca (USAL)
Repositorio:GREDOS. Repositorio Institucional de la Universidad de Salamanca
OAI Identifier:oai:gredos.usal.es:10366/160188
Acceso en línea:http://hdl.handle.net/10366/160188
Access Level:acceso abierto
Palabra clave:Ensemble Algorithm
Feature Ranking
Gradient Boosting
Machine Learning
Random Forest
Software development effort estimation
Descripción
Sumario:To provide a client with a high-quality product, software development requires a significant amount of time and effort. Accurate estimates and on-time delivery are requirements for the software industry. The proper effort, resources, time, and schedule needed to complete a software project on a tight budget are estimated by software development effort estimation. To achieve high levels of accuracy and effectiveness while using fewer resources, project managers are improving their use of a model created to evaluate software development efforts properly as a decision-support system. As a result, this paper proposed that a novel model capable of determining precise accuracy of global and large-scale software products be developed with practical efforts. The primary goal of this paper is to develop and apply a practical ensemble approach for predicting software development effort. There are two parts to this study: the first phase uses machine learning models to extract the most useful features from previous studies. The development effort is calculated in the second phase using an advanced ensemble method based on the components of the first phase. The performance of the developed model outperformed the existing models after a controlled experiment was conducted to develop an ensemble model, evaluate it, and tune its parameters.