Normal distribution

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.

The Normal or Gaussian distribution is the bell-shaped probability distribution that is often found in natural phenomena such as weights and heights. Some forms of statistical anlaysis may assume that data is Normally distributed. Also some algorithms may use the Normal distribution as part of theor processing, for example Gaussian filter for smoothing an image. Normal distributions are very common because they arise when large numbers of nearly independent small items are added tgeher (e.g. the effect of large numbers of genes and environmental effects). Even though they are common, it is important to remember that there are also many kinds of non-Normal distrubution for example power-law distributions that occur in much web and social network data.

Used on Chap. 7: pages 131, 140, 141, 142, 147; Chap. 10: pages 206, 207; Chap. 12: page 251; Chap. 14: page 335

Also known as Gaussian distribution, Normally distributed