Filtr Pareto w procesie syntezy klasyfikatorów wieloetykietowych w algorytmach wspomagania diagnostyki medycznej
Journal Title: Computer Science and Mathematical Modelling - Year 2015, Vol 0, Issue 1
Abstract
W pracy przedstawiono możliwość wykorzystania metod optymalizacji wielokryterialnej w procesie fuzji klasyfikatorów prostych w bardziej precyzyjne i wiarygodne klasyfikatory złożone. Przedstawiono proste (jednoetykietowe) klasyfikatory w postaci komitetów klasyfikatorów, pozwalające uzyskiwać klasyfikatory złożone o lepszych własnościach.
Authors and Affiliations
Andrzej Ameljańczyk
Keywords
Related Articles
- EP ID EP77432
- DOI 10.5604/15084183.1197401
- Views 94
- Downloads 0
Qualification of variables’ priorities for some decision functions
The following thesis describes two methods of finding variables’ priorities for some specific logical functions (without any negated variables). The first method is based on Hamming distance and requires a review of exam...
Foundations for Spread Page: review of existing concepts, solutions, technologies capabile of improving effectiveness of conveying knowledge
Spread Page is our code name for a new, more efficient way of conveying technical information and scientific knowledge – freed form the text-centered mindset and focused on graphical, interactive, multidimensional repres...
Model of attrition process in the presence of decoys
CART – CREATIVE APPLICATION TO REMEDY TRAFFIC
CART is a simulation software assisting decision-making process, aiming at enhancing the throughput of transport infrastructure and “fluidity” of vehicle traffic in urban agglomerations. Thanks to innovative solutions co...
Test methods for selected value-added services in wide area networks
In this paper selected value-added services used in wide area networks are characterized, as well as there are also discussed methods for formal description of the selected value-added services to their testing and quali...