Random Forest Based User Story Effort Estimation Model for Scrum Projects Based on Supervised Learning

Abstract

This paper presents a novel approach to software effort estimation in Scrum projects using a user story-based model and supervised learning techniques. Effort estimation is a critical aspect of software engineering, with various projects requiring distinct feedback from users and customers. Our proposed model is designed to predict effort using three key attributes: story points, complexity, and priority. A synthetic dataset of 300 projects was generated to train a Random Forest Regressor, with performance evaluated using Mean Squared Error (MSE) and R-squared metrics. The model achieved a high R-squared value of approximately 0.94, indicating strong predictive accuracy. These results suggest that the model effectively estimates effort based on user stories and can reduce uncertainty in Agile project management. Future work will involve applying this approach to real-world data for further validation.