作者: Susana M Vieira , Joao P Carvalho , André S Fialho , Shane R Reti , Stan N Finkelstein
DOI: 10.1109/IFSA-NAFIPS.2013.6608408
关键词: Neural fuzzy 、 Bag-of-words model 、 Feature selection 、 Machine learning 、 Computer science 、 Data mining 、 Fuzzy logic 、 Word (computer architecture) 、 Fuzzy rule 、 Decision support system 、 Data modeling 、 Artificial intelligence
摘要: This paper presents a proposal for decision support system the prevention of Intensive Care Unit readmissions, that is being built based on data from MIMIC II database. The proposed fuses results two distinct classification approaches, one available numerical data, and other medical text annotations. former uses neural fuzzy models modeling in order to provide interpretable results. latter creation top-k edited word lists fingerprints. Model fusion accomplished using two-tiered rule architecture.