wavelet transform

Terms from Artificial Intelligence: humans at the heart of algorithms

Page numbers are for draft copy at present; they will be replaced with correct numbers when final book is formatted. Chapter numbers are correct and will not change now.

A wavelet transform decomposes a function or time series into a sum of products of wavelets from a collection. The collection is often defined mathematically using a kernal, which then mans thet thay have good properties. It is rather like a Fourier transform, as the wavelets typically opertae at different tmeoral or spatial scals, but whereas the sine/cos funcitons used in the Fourier transform extend over the etire temproal or spatial domain, wavelets are local.

Defined on page 322

Used on Chap. 8: page 155; Chap. 12: page 276; Chap. 14: pages 321, 322, 330, 333

Also known as wavelet transformation

Mexican hat' wavelet

A simple family of wavelets -- the Haar Wavelet