3_10_09_SupervisedLearning

# 3_10_09_SupervisedLearning - Spider This model develops...

This preview shows pages 1–11. Sign up to view the full content.

Spider!

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This model develops quasi-periodic columns, with variations due to random initial conditions.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Unsupervised learning is self-organization to maximize extraction of information from input.
How can we modify networks like these to learn how to perform tasks well?

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
One way is by having a supervisor tell the network whether its performance is good.
Another way is by having a supervisor tell the network what is the correct answer. Spider!

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
In learning models, the dynamics of firing are much faster than those of synaptic plasticity. Hence, a good approximation is The simplest rule following Hebb’s conjecture is v = w u τ w d w dt = v u In the simplest case, one uses this rule with w d w dt = v u
In Hebbian supervised learning without error correction, the output v is the correct answer given in samples (superscripts are not powers): For stability of weights, we add decay: τ w d w dt = v u w d w dt = 1 N S v m u m m = 1 N S w d w dt = −α w + 1 N S v m u m m = 1 N S

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Example 1: In a perceptron, a nonlinear map that classifies binary-vector inputs into one of two categories, the desired output is A perceptron only classifies inputs perfectly under the condition of linear separability. This condition is the existence of a hyperplane that divides the input space such that one of the portions corresponds to v = 1 and the other to v = -1.
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 06/08/2009 for the course BME 575L taught by Professor Grzywacz during the Spring '09 term at USC.

### Page1 / 36

3_10_09_SupervisedLearning - Spider This model develops...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online