In order not to miss potential predictors, prognostic researchers tend to gather an excessive amount of data, after which a smaller set of predictors is selected using statistical methods.
Anticipated advantages of the CART vs. simple additive linear and logistic models were less than expected in this particular application with a relatively small set of predictor variables.
The present findings are based on juxtaposing the results for the very simplest additive, algebraic, linear and logistic regression vs. the non-additive, non-algebraic CART models based on the relatively small set of predictor variables examined.
To differentiate a smaller set of statistically significant predictors from those with similar predictive ability we performed 10000 additional L20OCV iterations which enabled us to choose three best FLP (each is based on 10 genes; see Table 4).
We first built a large number of RFs to identify and rank TFBSs of importance; and then supplied the resultant TFBSs as a relatively smaller set of predictor variables to CART for classification, using step-wise forward selection procedure.
The Bayes quadratic model required the smallest set of predictor variables (only three).
The Bayes quadratic model required the smallest set of predictor variables (only three: oxygen extraction, oxygen delivery and use of cardiac inotropic drugs after the operation) and provided very interesting results, which were similar or better than those obtained with the Bayes linear or logistic regression models.
A small set multivariate predictor may have important applications in the early detection of neoplastic transformation in populations at high risk for HCC, such as hereditary haemochromatosis patients [18].
Essentially, this is a feature selection or a gene selection task [ 28, 32, 36, 37], where the goal is to model the target response Y k with an optimal small set of important predictor variables, i.e., a subset of columns of the X− k matrix.
At least for binary prediction of dyslipidemia from waist-to-hip ratio and body mass index in the context of the relatively small set of other predictor variables examined, the simple additive logistic models obtained in previous studies were about as effective as the more comprehensive statistical models investigated here.
