Proactive Code Churn Management with Version Control Systems and Artificial Intelligence

Managing code churn, the frequent and significant modification of code within a project, remains a persistent challenge in software development. This research paper addresses this issue by presenting a novel approach to proactive code churn management through the integration of version control systems (VCS) with artificial intelligence (AI). Specifically, a hybrid architecture of neural networks is employed to predict the code churn ratio, the proportion of code lines changed relative to the total number of code lines in a project. The proposed model achieves a Mean Absolute Error (MAE) of 0.0213, demonstrating high accuracy in forecasting code churn. By combining the strengths of convolutional layers for feature extraction and recurrent layers for sequence modeling, the hybrid neural network architecture offers a robust and effective framework for churn prediction. This research contributes to enhancing software development processes by proactively addressing code churn challenges.