A Novel Artificial Neural Network Approach for Predicting Development Effort Using Use Case Points

Accurate software development effort estimation is vital to prevent budget overruns, missed deadlines, and project failures. This study introduces an Artificial Neural Network (ANN) model that leverages the Use Case Points (UCP) approach to estimate effort during early development stages. By incorporating project size, productivity, and a novel complexity metric, the ANN model outperforms traditional regression and UCP methods, achieving 8% and 50% lower Mean Magnitude of Error Relative to the Estimate (MMER), respectively. The results highlight the ANN model's robustness and accuracy as a reliable tool for early effort estimation.