Tom M' Mitchell

1
Machine Learning
Tom M. Mitchell Carnegie Mellon University March
2003 For more, see Machine Learning, Tom
Mitchell, McGraw Hill, 1997.
2
Learning to Predict Emergency C-Sections
9714 patient records, each with 215 features
3
Artificial Neural Network for Recognizing Speech
4
(No Transcript)
5
(No Transcript)
6
Train Software to Decode Cognitive States from
fMRI data
Examining a verb
7
Function Approximation Problem

• Problem setting
• set of instances X
• target function to be learned f X ! Y
• candidate hypotheses h X ! Y
• Input
• Training examples lt xi, yi gt
• Output
• hypothesis h that is best estimate of f

8
Function Approximation

• Many algorithms for different problems
• Decision trees
• Artificial neural networks
• Linear regression, Logistic regression
• k Nearest Neighbor
• naïve Bayes classifier
• Bayesian networks
• Support Vector Machines

9
Bayes Rule
Which is shorthand for
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
For code, see www.cs.cmu.edu/tom/mlbook.html
click on Software and Data
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
How can we implement this if the ai are
continuous-valued attributes?
26
Also called Gaussian distribution
27
Gaussian
Assume P(aivj) follows Gaussian distribution,
use training data to estimate its mean and
variance