Iterative Feature Elimination Method Using Artificial Neural Network for Software Effort Estimation
Pranay Tandon, Ugrasen Suman · International Journal of Engineering Trends and Technology · 2024
Effort estimation is one of the critical tasks for any software development team because estimation is the key to planning the software development life cycle activities with proper timeline and cost. On-time and quality delivery is most important to build customer trust and certainty. There are many features to be considered while estimating the efforts, but removing the weak features and finding the set of the strongest features for any estimation process is difficult. Deep learning is the most popular prediction technique for effort estimation because of its capacity to adapt and be accurate on different types of datasets. Artificial Neural Network is best suited to deep learning techniques for predicting effort, per industrial research. In this paper, a novel model based on artificial neural networks and an iterative feature elimination-based method has been proposed to estimate the efforts. With ranking features, the proposed method can find the optimized set of features to be used in the model and final efforts. COCOMO NASA 2 dataset is used to find the results.