15 comments by 2 or more people

1. i hates you.

   I'm still learning the module for the first time. The problem isn't with any conceptual difficulty but remembering lots of facts, algorithms and proofs.

   20 May 2005, 23:24

2. Yeh. Welcome to what I excel at. Don't worry, I'll get screwed over on Tuesday – 2.5 days to learn Magnetic Properties of Solids Oleg and NMR from scratch.

   20 May 2005, 23:58

3. Mathew Mannion

   You suck

   link

   20 May 2005, 23:59

4. See Dan's entry. I've already contributed a bit :-P

   21 May 2005, 00:04

5. do you know of various examples for different things? specificially all the questions when it asks you to list attributes and whatnot and illustrate with examples.

   21 May 2005, 00:09

6. erm. I might do. If I knew exactly what you meant.

   21 May 2005, 00:14

7. well all those nice easy questions where it asks you to list properties and then give examples. environment properties, fundamental agent types etc.

   21 May 2005, 00:18

8. are right. Agent properties = PEAS.
   Environment properties:
   Deterministic/Stochastic/Strategic
   Episodic/Sequential
   Static/Dynamic
   Fully/Partially Observable
   Discrete/Continuous.

   Not so great on examples but I'll try one: environment properties for chess:
   Strategic, Sequential, Static, Fully Observable, Discrete.

   21 May 2005, 00:26

9. Agent types:
   Simple relfex agent
   Reflex with state
   Goal based
   Utility based.
   I can draw the diagrams and stuff…

   21 May 2005, 00:27

10. with agent types can you say some examples? I can do the explanations and diagrams but i'm rubbish at thinking up examples in exams.

    21 May 2005, 00:28

11. Ok, well to continue the theme I'd say that a chess-playing agent (or anything involved in adversarial search for that matter) would be utility based. Know the goal is useless if something is trying to stop you getting there and you don't know what move is most beneficial.

    Goal based…hmm…how about the robot thingy we did in CS118? Before we programmed it to avoid repetitions it was simply goal based wasn't it – knew nothing but what it was trying to get to.

    The others I can't really think of right now. I'll let you know if I come up with something

    21 May 2005, 00:37

12. In just over 12 hours I have to sit a paper consisting of twelve 'questions'. 'Questions' = words. Single, evil words. Eg: "Transcendental". "Normativity". "Qualia". "Sublime". I shall be choosing one word and writing a 'philosophical essay' on it (once I have spent half an hour making up a definition for it and trying to figure out at what point I was supposed to have studied it) for three long hours.

    WIN!

    21 May 2005, 01:16

13. sounds fun. i'm serious here. such restrictions allow some amazing creativity.

    21 May 2005, 01:19

14. Ouch

    21 May 2005, 01:19

15. Perhaps, if 'philosophical essay' didn't mean "you should know the correct terminology / history of the topic" blah blah blah… Also it's hard to be creative with words like "qualia", when one can only vaguely remember which branch of philosophy it's supposed to be a relevant part of!

    *Sigh. All will be well. I can be philosophical and stuff… right?

    21 May 2005, 01:22