Metaheuristic design pattern: Interactive solution presentation
Year: 2014
Type: conference-paper
Abstract: In interactive metaheuristic search, the human helps to steer the trajectory of search by providing qualitative evaluations of solution individuals in the population. Given that much meta-heuristic search is population-based, it is challenging to design the presentation of solutions such that the human can provide effective qualitative evaluation [1]. Naively presenting each solution individual in a large population at each generation causes evaluation fatigue and a subsequent non-linearity of user focus making search trajectory inconsistent and ineffective [2]. Interactive search relies either on solely user-provided qualitative evaluation or a combination of user qualitative evaluation and machine-based quantitative fitness measures. User qualitative evaluation is often “multi-subjective ” in that many fitness concerns are simultaneously evaluated [3]. Some evaluative concerns may be explicitly articulated by the user, although others
Authors: Mark R. N. Shackelford, Christopher L. Simons
Language: English
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7020930442
