Feature Selection with Multivariate Symmetrical Uncertainty to predict Dengue Cases using Deep Learning

Marcos Ortega, Santiago Gómez, Fredy Ramı́rez, Héctor Estigarribia

Resumo


Dengue is a viral disease transmitted by the Aedes aegypti female mosquito, affectingvast areas of the world. In the last 50 years, its incidence in Paraguay has increased,accompanying the persistent migration into the cities [1]. Approximately 80 million cases appear every year in more than 100 countries, and about 2.5 billion people live in countries with endemic dengue. Paraguay is part of this list of countries, as one of the most affected by the disease [3]. Since the appearance of dengue in Paraguayan territory there has been a scalar increase in policies, strategies and public health services that prevent and combat the outbreaks. Despite all these efforts, large epidemics were recorded in the 1988-1989; 1999-2000; 2006-2007 and 2012-2013 periods [1]; and currently there are many cases of the disease in the country. In this work we propose a model to forecast the number of probable dengue cases using two techniques in tandem. First, we implement a novel technique for feature selection using Multivariate Symmetric Uncertainty (MSU) [2], which we employ to compare feature  sets. Secondly, the selected feature sets are used to feed a deep learning neural network.[...]


Texto completo:

PDF

Apontamentos

  • Não há apontamentos.


SBMAC - Sociedade de Matemática Aplicada e Computacional
Edifício Medical Center - Rua Maestro João Seppe, nº. 900, 16º. andar - Sala 163 | São Carlos/SP - CEP: 13561-120
 


Normas para publicação | Contato