Memetic Evolutionary Multi-Objective Neural Network Classifier to Predict Graft Survival in Liver Transplant Patients

Áreas de investigación:

• Methodology - Machine learning

Año:

2011

Tipo de publicación:

Artículo en conferencia

Palabras clave:

Artificial neural networks, Generalized radial basis functions, Liver transplantation, Multi-objective evolutionary algorithm

Autores:

• Cruz-Ramírez, Manuel
• Fernández, Juan Carlos
• Fernandez-Navarro, Francisco
• Briceño, Javier
• de la Mata, Manuel
• Hervás-Martínez, César

Título del libro:

Proceedings of the 13th annual conference companion on Genetic and evolutionary computation (GECCO2011)

Páginas:

479-486

Organización:

Dublin, Ireland

Mes:

12-16 July

ISBN:

978-1-4503-0690-4

BibTex:

@conference{Gecco2011, author = "Manuel Cruz-Ram{\'i}rez and Juan Carlos Fern{\'a}ndez and Francisco Fernandez-Navarro and Javier Brice{\~n}o and de la Mata, Manuel and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "In liver transplantation, matching donor and recipient is a problem that can be solved using machine learning techniques. In this paper we consider a liver transplant dataset obtained from eleven Spanish hospitals, including the patient survival or the rejection in liver transplantation one year after it. To tackle this problem, we use a multi-objective evolutionary algorithm for training generalized radial basis functions neural networks. The obtained models provided medical experts with a mathematical value to predict survival rates allowing them to come up with a right decision according to the principles of justice, efficiency and equity.", booktitle = "Proceedings of the 13th annual conference companion on Genetic and evolutionary computation (GECCO2011)", doi = "10.1145/2001858.2002037", isbn = "978-1-4503-0690-4", keywords = "Artificial neural networks, Generalized radial basis functions, Liver transplantation, Multi-objective evolutionary algorithm", month = "12-16 July", organization = "Dublin, Ireland", pages = "479-486", title = "{M}emetic {E}volutionary {M}ulti-{O}bjective {N}eural {N}etwork {C}lassifier to {P}redict {G}raft {S}urvival in {L}iver {T}ransplant {P}atients", url = "http://dx.doi.org/10.1145/2001858.2002037", year = "2011", }

Abstract:

In liver transplantation, matching donor and recipient is a problem that can be solved using machine learning techniques. In this paper we consider a liver transplant dataset obtained from eleven Spanish hospitals, including the patient survival or the rejection in liver transplantation one year after it. To tackle this problem, we use a multi-objective evolutionary algorithm for training generalized radial basis functions neural networks. The obtained models provided medical experts with a mathematical value to predict survival rates allowing them to come up with a right decision according to the principles of justice, efficiency and equity.