Integrating machine learning models with continuous integration and continuous delivery (CI/CD) pipelines for a learning-driven approach to software engineering

Abstract

The report addresses the topic of integrating Machine Learning (ML) models into Continuous Integration/Continuous Delivery (CI/CD) pipes to improve automation and enable stringent model validation and facile deployment procedures. Accuracy, recall, and F1 score are key performance indicators (KPI) used to ensure the best operation of the model. This paper highlights the need to perform error detection, detect data drift and overfitting, and the need to constantly retrain. It also analyses the effects of the volume of data, vitality of features, and perception stability. The conclusions show that the integration of ML models in CI/CD pipelines enhances the effectiveness of models to better support effective decision-making and positive business results.