Objective
In this 3-day course we elaborate all available configurations in SAM and train on how to apply standard validation techniques to select optimal configuration settings. We will also elaborate the short-term forecast feature in SAM.
During the course we set up configurations from scratch for several stocks, and participants are encouraged to bring their own dataset that can be addressed in plenum.
Day 1: Introduction to SAM
- Basic theory behind SAM
- Introducing all configurations available in SAM
- Introducing standard validation: AIC, one-step-ahead residuals, process residuals, retro, cross validation, leave-out, jitter analysis and simulation study
Day 2:
- Propose configurations for two different data sets
- Introduce forecast feature in SAM
Day 3:
- Propose configurations for two different data sets
- Participants are encouraged to bring their own dataset that can be addressed in plenum
Note that program in not finalized, and changes may occur.
Level
Participants to this course are expected to be students or scientists who would normally participate in working groups. No formal training in statistics will be needed for participating, and programming will be kept at a minimum.