Machine learning(4): Instance based learning

Numeric Attributes:

For example:

Nominal Attributes

When the value of i is same as the value of j, the nominal attribute is 1 and other situtaion is 0.

Distance Between Two Examples

k-Nearest Neighbour

When the y is equal to z, the value of Q(y,z) is 1. If not, the value of Q is 0.

Distance Weighted Nearest Neighbour

Addding the distance is to calculate the weighted of nearest neighbour. The distance affects the value of Q.

Advantages of IBL

1. Very good method of learning representations of complex functions.
2. Simple representation .
3. Fast learning .