multi-linear regression

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.

Multi-objective optimisation is used when there are several criteria, all of which one would wish to optimise. For example, a company may want to optimise profit, welfare for its workers and environmental impact. Typically there is not a way to simultaneously optimise all the criteria so there has to be a trade-off between them. Multi-objective optimisation is hence far more difficult than simple single-criterion optimisation. One solution is to weight the criteria and optmise the weighted sum. Other approaches attempt to find some or all Pareto-optimal solutions; that is solutions from which there is no way to improve all citeria.

Used on Chap. 7: page 131