Has anyone taken this course with him? His reviews on RMP aren't that great, and my course selection is kind of getting short so I can't wait until Mori teaches it.
mmmm? Care to elaborate? I love dealing with AI, but I'm not sure if I'm capable of it. Hadley doesn't even have a website setup for CMPT320, and for something like this, that's a really bad sign.
EDIT: What about cmpt300 with Bradley Bart? His reviews is quite good, but it's a morning class.
310 with Hadley is suicide! i took it the semester AFTER taking 320 with him so i knew what to expect... but it was much worse then expected
he makes u write CODE on midterm and final you lose 3% of ur mark for any syntactical error (WTF!!!!) his marking is brutal, you can have the right answer ... heck for my midterm i had all the right answers but you lose bullshit marks all over the place till u end up with some shit on a stick grade
cmpt320 aint bad with him ... all u need to do is memorize the notes for exams but 310 is memorize notes and have perfect English and be a fucking compiler although the assignments are easy in 310 with Hadley
he uses email for everything, he dosnt post any notes and he talks like a fag and makes that retarded hand gesture every time hes trying to sound smart...
also keep in mind his phd is in philosophy not computing
I have friends taking 320 with him this semester and they mentioned that he's quite intellectual. I thought that would make up his crappy teaching skills.
How's CMPT310 like in general though? I guess I'll wait for Mori then.
I think your friends meant "intellectual" in a derogatory way.
Intellectual probably means stuck-up jackass that expects way too much from the student and is anal on every little thing. Because he's so perfect himself, he expects no less from the students.
CMPT 310 - Bob Hadley - [email]hadley@sfu.ca[/email] Marking Assignment 1 15% Midterm Exam 25% Project 25% Final Exam 35%
and to give u an idea of his notes....
Two main goals ○ Practical programs that do tasks requiring real intelligence (an Engineering Approach) ○ Understand how humans solve difficult problems § To match or surpass human intelligence as ultimate goal (a Scientific Approach)
Benefits of AI for understanding human thought processes a. AI can supply concepts for theories of the mind § Procedure § Loops § Concept of a program b. Building computer models of human mind enables cognitive theories to be made precise and testable. c. AI programs allow us to put upper bounds on Resources Needed § Computational complexity d. Some general principals are at times discovered that enable us to be better at problem solving
Different Approaches to building AI systems ○ Most AI programs are domain specific (Narrow in scope) § These are often the most useful AI programs § Ex. Deep Blue ( play chess ) § Often such programs are "brittle". ○ Many of the "interesting" AI programs are more general and more flexible and employ deep strategies. § Ex. "Automated mathematician" (D. Lenant) □ Discovered important conjectures (does not prove theories) § Ex. "General problem solver" (GPS) (Newell and Simon)
Major Areas of AI ○ Expert Systems § Ex: □ Mycin by Shortliff ® bacterial analysis - better then family doctor or even experts □ Prospector by Duda, Hart, Reboh ® Geological advice for mining □ Dendral ® Analysis of spectrograms ® Create chemical formula and molecular diagram ○ Neural Networks ( connectionism) Neuron - brain cells Loosely based on how the human brain works § Can learn to recognize patterns □ Visual patterns □ Syntactic patterns § Learn concepts or something like concepts § Association of ideas ○ Automated Reasoning § Theorem proving in logic or mathematics § Common sense logic have been explored □ "default" reasoning § Game playing □ Methods for "board" games ® Chess ® Checkers ® Etc.
Comments
NEVER!!!!!!!!!
EVER!!!!!!
STOP THINKING ABOUT IT!!!
NO!! I SAID NO!!!
STOP!!!
EDIT: What about cmpt300 with Bradley Bart? His reviews is quite good, but it's a morning class.
i took it the semester AFTER taking 320 with him so i knew what to expect...
but it was much worse then expected
he makes u write CODE on midterm and final
you lose 3% of ur mark for any syntactical error (WTF!!!!)
his marking is brutal, you can have the right answer ... heck for my midterm i had all the right answers but you lose bullshit marks all over the place till u end up with some shit on a stick grade
cmpt320 aint bad with him ... all u need to do is memorize the notes for exams
but 310 is memorize notes and have perfect English and be a fucking compiler
although the assignments are easy in 310 with Hadley
he uses email for everything, he dosnt post any notes and he talks like a fag and makes that retarded hand gesture every time hes trying to sound smart...
also keep in mind his phd is in philosophy not computing
How's CMPT310 like in general though? I guess I'll wait for Mori then.
Intellectual probably means stuck-up jackass that expects way too much from the student and is anal on every little thing. Because he's so perfect himself, he expects no less from the students.
Marking
Assignment 1 15%
Midterm Exam 25%
Project 25%
Final Exam 35%
and to give u an idea of his notes....
Two main goals
○ Practical programs that do tasks requiring real intelligence
(an Engineering Approach)
○ Understand how humans solve difficult problems
§ To match or surpass human intelligence as ultimate goal
(a Scientific Approach)
Benefits of AI for understanding human thought processes
a. AI can supply concepts for theories of the mind
§ Procedure
§ Loops
§ Concept of a program
b. Building computer models of human mind enables cognitive theories to be made precise and testable.
c. AI programs allow us to put upper bounds on Resources Needed
§ Computational complexity
d. Some general principals are at times discovered that enable us to be better at problem solving
Different Approaches to building AI systems
○ Most AI programs are domain specific (Narrow in scope)
§ These are often the most useful AI programs
§ Ex. Deep Blue ( play chess )
§ Often such programs are "brittle".
○ Many of the "interesting" AI programs are more general and more flexible and employ deep strategies.
§ Ex. "Automated mathematician" (D. Lenant)
□ Discovered important conjectures (does not prove theories)
§ Ex. "General problem solver" (GPS) (Newell and Simon)
Major Areas of AI
○ Expert Systems
§ Ex:
□ Mycin by Shortliff
® bacterial analysis - better then family doctor or even experts
□ Prospector by Duda, Hart, Reboh
® Geological advice for mining
□ Dendral
® Analysis of spectrograms
® Create chemical formula and molecular diagram
○ Neural Networks ( connectionism)
Neuron - brain cells
Loosely based on how the human brain works
§ Can learn to recognize patterns
□ Visual patterns
□ Syntactic patterns
§ Learn concepts or something like concepts
§ Association of ideas
○ Automated Reasoning
§ Theorem proving in logic or mathematics
§ Common sense logic have been explored
□ "default" reasoning
§ Game playing
□ Methods for "board" games
® Chess
® Checkers
® Etc.