|
LO-1
|
Upon successful completion of this course, students will gain the competence to integrate DevOps principles into data science projects by understanding the machine learning lifecycle and MLOps maturity levels; ensure the reproducibility of machine learning models by performing experiment tracking, model, and data versioning using tools like MLflow and DVC; package models using container architectures and deploy them for batch or online serving via continuous integration and deployment (CI/CD) pipelines; track the performance of models in production to detect situations such as data drift and concept drift; and simultaneously design real-time data streaming systems with message queues, stream processing, and windowing techniques using Apache Kafka, integrating these systems into end-to-end machine learning projects with the help of workflow orchestration tools. |