Transfer Learning Across Correlated Instruments

Use when a new or thinly traded instrument has too little history to fit a model on its own and a liquid co-moving instrument exists; fits on the source and adapts, rather than fitting noise on the target.