Projects
!
Project partners
!
Assignment: apply a machine learning technique to some non-trivial data set
!
I would like people to form groups of 2 or 3 preferably to do the project
!
Can be technique discussed in class or one that you look up on your own
CSE 730: Artificial Intelligence 2 Advanced Topics
Prof. Eric Fosler-Lussier TR 2-3:18 Dreese 264 Autumn 2007
How do you say that name?
!
Prof. Fosler-Lussier (faaz-ler loo-see-er)
!
But there are at least 2 variants of Fosler and 3 variants for Lussier w
Logical reasoning Week 2
Probability theory Bayes Nets
!
Here, the variables are logic sentences
Mary loves John S1: loves(M,J)
Betty loves John
S2: loves(B,J)
John loves John
S3: loves(J,J)
Q: Does everyone love John?
S1^S2^S3=> ! x loves(x,J)
Last week
Reminder from last time Week 3
Bayesian Inference Modeling time series Alarm example ! Compute: P(j,m,a,~b,~e)= P(j|a)P(m|a)P(a|~b,~e) P(~b) P(~e) = .9 * .7 * .001 * .999 * .998 = 0.000628
!
P(b) 0.001
B A
E
P(b) 0.002 B T T F F E T F T F P(a) .95 .94 .29
What can be learned? Week 4
Learning Chapter 18, 20
! !
Classifications:
! ! !
Identify hand-written digits Filter mail into spam/not spam Find the face in a photo
Actions:
! ! !
Robot balances upright on two legs Autopilot flies level Keep vehicle in lan
What went right Week 5
Quiz 1 Redux Learning Vision Chapter 20, 24
! !
Almost everyone got
!
Inference cloud problem (Problem 4) Figuring out how to break down probabilities into components Probability distributions
Many people got
! !
Quiz 1
!
What went
Vision as perception Week 6
Vision
!
We can think of our percepts as arising from some view of the world W
P = g(W)
!
If g is graphics, perhaps we can do vision as inverse graphics?
W=g-1(P) ?
!
Problem: Many worlds are possible
Perception
!
Vision as per
Audition
Vision: use eyes/camera and extract perceptual information ! Audition: use ears/microphone and extract perceptual information ! A little experiment
!
Fundamental frequency @ 200 Hz
Audition tasks
! !
Adding in a harmonic @ 400Hz
Computational Aud
Probabalistic CFGs Week 9
Language Processing
!
!
If W is a word sequence, and T is a tree:
P (W ) = " P (W , T ) = " P (W | T ) P (T )
T T
Why multiple trees? There can be multiple ! parses of a sentence. ! How to calculate P(T)? If T is headed by rule R