In a collaborative project where a team of data scientists aims to track and compare the performance of various machine learning experiments, which MLflow component is most suitable for this purpose?