This repository is the coursework of model AIN when I studied at King's College London from 2019 to 2020 In the pacman file, there is a .py file called MDPAgents.py contains the whole Markov Decision Process program, all the other files are not my work. They are provided by my tutor, and they were developed at UC Berkeley.
(a) you can test the code in smallGrid by running:
python pacman.py -q -n 10 -p MDPAgent -l smallGrid
-l is shorthand for -layout. -p is shorthand for -pacman. -q runs the game without the interface (making it faster).
(b) in mediumClassic by running:
python pacman.py -q -n 10 -p MDPAgent -l mediumClassic
The -n 10 runs ten games in a row.
the feedback of this coursework is shown as follow:
- Coding
Code style is good (but with some big blocks of code that are confusing – for in for in for…). Comments are few and quite unhelpful.
- Report
The description is really good! Congrats! The evaluation should be expanded, with each component to be tested and evaluated.