作者: Enrique Romero Merino , Alfredo Vellido Alcacena , Juan Carlos Ruiz Rodríguez , Vicent Ribas Ripoll
DOI:
关键词:
摘要: In this paper, we describe a novel kernel for multinomial distributions, namely the Quotient Basis Kernel (QBK), which is based on suitable reparametrization of input space through algebraic ge- ometry and statistics. The QBK used here data transformation prior to classification in medical problem concerning prediction mortality patients suffering severe sepsis. This common clinical syn- drome, often treated at Intensive Care Unit (ICU) time-critical context. Mortality results with Support Vector Machines using compare favorably those obtained alternative kernels standard procedures.