Oct 23, 2006
-------------

- Learning methodology
	- training set and test set
	- error curves as function of training set size
	- knowing when to stop learning
		- overfitting boundary
	- avoid "peeking"!

- Twists to the tale
	- what to do with continuous attributes?
		- binning strategies

- Troublesome datasets for decision trees
	- parity function
	- majority function
	- m-of-n functions

- Problems with decision tree learning
	- fragmentation
	- duplication of subtrees
		- arises because of greedy
		  algorithm

- Next topic: neural networks 
	- input nodes, hidden nodes, output nodes
	- autonomous car driving example
		- at CMU

- Introduction to neural networks
	- perceptrons
	- layers of perceptrons => neural network

- Biological motivation and metaphor
	- axon
	- dendrites
	- processes taking place inside a neuron

- Models of artificial neurons (perceptrons)
	- weighting different inputs;
	- summing; and
	- thresholding

- What are neurons capable of modeling
	- linear planes in hyperspace
		- e.g., AND in 2D space
		- e.g., OR in 2D space

- Realization of classical gates in neurons
	- weights for AND gate
	- weights for OR gate

- How can we do an XOR?
	- not possible with one neuron!
	- try to do it with two layers of neurons!