Next: Radial Basis Functions
Up: Instance-Based Learning
Previous: Locally Weighted Regression
- broad range of methods for distance weighting the training examples
- range of methods for locally approximating target functions
- function is usually constant, linear, or quadratic because (1) cost
of fitting more complex functions is too high and (2) simple
approximations model the target function well over a sufficiently
small subregion
Patricia Jean Riddle
Wed Jun 23 13:06:34 NZST 1999