Searching for Risk in Large Complex Spaces

Autor:	Kester Clegg, Rob Alexander
Rok vydání:	2014
Předmět:	Mutation operator Focus (computing) business.industry Computer science Complex system Space (commercial competition) Air traffic control Machine learning computer.software_genre Work (electrical) Artificial intelligence Heuristics business Risk assessment computer
Zdroj:	Applications of Evolutionary Computation ISBN: 9783662455227 EvoApplications
DOI:	10.1007/978-3-662-45523-4_61
Popis:	ASHiCS (Automating the Search for Hazards in Complex Systems) uses evolutionary search on air traffic control simulations to find scenario configurations that generate high risk for a given air sector. Weighted heuristics are able to focus on specific events, flight paths or aircraft so that the search can effectively target incidents of interest. We describe how work on the characterization of our solution space suggests that destructive mutation operators perform badly in sensitive, high dimensional spaces. Finally, our work raises some issues about using collective risk assessment to discover significant safety events and whether the results are useful to safety analysts.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::ff5d0be0d66d58a359815742ef126e6d https://doi.org/10.1007/978-3-662-45523-4_61 Zobrazit plný text záznamu