Tutorial Agenda (Tentative)
Time | Content | Readings | Material |
---|---|---|---|
15 mins | Introduction to SpeedyIBL & SpeedyIBL installation | Reading | Note Slides |
20 mins | Demo of IBL model for repeated Binary Choice task | Reading | Code |
20 mins | Hands-on exercise: Iowa Gambling Task (IGT) | Exercise Demo |
Code |
10 mins | Sequential decision-making tasks in Gridworld environment | Demo | |
30 mins | Demo of IBL model for goal-seeking task in Gridworld environment (Scenario 1): there is only one target and consuming the target terminates the game. | Demo | Code 1 Code 2 Code 3 |
10-min Break | |||
25 mins | Hands-on exercise: Goal-seeking task in Gridworld environment (Scenario 2): there are two targets, one with a high value and the other with a low value. The goal is to find the higher value target. | Exercise Demo |
Code 1 Code 2 Code 3 |
25 mins | Hands-on exercise: Goal-seeking task in Gridworld environment (Scenario 3): there are two targets, and the goal is to collect all the targets. | Demo | Code 1 Code 2 Code 3 |
30 mins | Hands-on Exercise: SpeedyIBL for Atari game environments developed by Open Gym AI task (Atari games): Ms. Pacman | Gym Atari Exercise |
Atari ROMS Code |