Hi all:
Thanks for all the hard work this semester. I hope you had fun. Some of this is similar to the mid-course survey, but there is a bunch added to the end.
How was the speed of the course?
Way too slow
Sort-of too slow
Just about right
Sort-of too fast
Way too fast
Have the written homeworks been reasonable?
Too easy: they didn't help me learn anything
About right
Too hard: they took a ridiculous amount of time
Were the projects reasonable?
Way too easy
Sort-of too easy
About right
Sort-of too hard
Way too hard
Listed below are a bunch of topics we covered in class. For each, please mark two things: (1) how interesting you thought the topic was; and (2) whether you think the topic was covered adequetly. For interesting-ness, "1" means not-at-all interesting, "2" means moderately interesting, "3" means very interesting. For coverage, "1" means that you'd like to see more, "2" means it was covered just right, "3" means you would like to have seen less. You're free to leave some options blank if you don't have an opinion.
TOPIC
INTEREST
COVERAGE
Search
1
2
3
1
2
3
Constraint Satisfaction
1
2
3
1
2
3
Game Playing
1
2
3
1
2
3
Markov Decision Processes
1
2
3
1
2
3
Reinforcement Learning
1
2
3
1
2
3
Graphical Models
1
2
3
1
2
3
Sampling
1
2
3
1
2
3
HMMs
1
2
3
1
2
3
Speech
1
2
3
1
2
3
Machine Translation
1
2
3
1
2
3
Machine Learning
1
2
3
1
2
3
There are many topics that are covered in a "standard" intro to AI class that we did not cover. These include: first order logic, theorem proving, knowledge representation, perception, robotics, planning, and AI philosophy. Are there any topics here (or anything else!) that you really wish had been covered (keeping in mind that it would mean something else would probably have to go):
Feedback on the projects would be very useful. For each project, please select how interesting you thought it was (as above), how difficult you thought it was (1 means too easy, 3 means too hard), and how well you thought it fit in the class (1 means not at all, 2 means fit reasonably well, 3 means very good fit).
PROJECT
INTEREST
DIFFICULTY
FIT
Project 1: Pacman Search
1
2
3
1
2
3
1
2
3
Project 2: Pacman Multiplayer
1
2
3
1
2
3
1
2
3
Project 3: Reinforcement Learning
1
2
3
1
2
3
1
2
3
Project 4: Static Ghostbusters
1
2
3
1
2
3
1
2
3
Project 5: Dynamic Ghostbusters
1
2
3
1
2
3
1
2
3
AI contains many subdisciplines. For some of these, we have existing classes (natural language processing, machine learning, computer vision, robotics), but there are lots of others. Are there any AI-related courses that you really wish were offered here at the U? Any and all suggestions are welcome; feel free to specify level (ie, 3000 level, 4000 level, 8000 level, etc.). Here is a list (written in white so you have to highlight it to see) of some ideas for such courses... it's in white, though, so you think about it a bit yourself before looking at our cheat sheet. [
information retrieval, data machine, web search, agent-based learning, reinforcement learning, computational biology, computer music, ...
]
Finally, any other comments are more than welcome:
Don't forget to