In optimisation and forms of machine learniong the fitness function (also known as objective function or utility function), measures how good a solution or state is. For example, the fitness function for an image reconstruction may be measured as the sum of the squares of difference between the pixel values of the target and generated images. Note that the choice of fitness function has important impact on efficiency, for example, if it has hard thresholds this may make incemental learning hard. The fitness function can also be a source of bias, for example, if you choose based on popularity you may effectively embody societal discrimination.
Used on pages 172, 179, 506
Also known as objective function