machine learning - What is multiobjective clustering? -

i don't understand multiobjective clustering using multiple variables clustering or what? know stack overflow might not best kind of questions, i've asked on other website , did not got response.

multiobjective optimization in general means have multiple criterions interested in, cannot converted comparable. example consider problem when try have fast model , accurate one. time measured in s, accuracy in %. how compare (1s, 90%) , (10days, 92%)? 1 better? in general there no answer. people - pareto front, test k models , selec m <= k of them such that, none of them "beaten" else. example if add (1s, 91%) previous example, pareto front {(1s, 91%), (10days, 92%)} (as (1s, 90%) < (1s, 91%), , remaining ones impossible compare).

and can apply same problem in clustering setting. example want build model fast classify new instances, minimizes avg. distance inside each cluster, , not put each cluster many special instances labeled x. again models (clusterings) characterized 3, not comparable, measures, , in multiobjective clustering try deal these problems (like example finding pareto front of such clusterings).