Watch this video on YouTube
Your team is working on optimizing Spark performance for a large-scale machine learning project. What technique involves storing intermediate data in memory to speed up iterative algorithms and reduce disk I/O?