partDSA: deletion/substitution/addition algorithm for partitioning the covariate space in prediction

The partDSA package (Molinaro, Lostritto, and Weston 2009) provides a novel recursive partitioning tool for prediction when numerous variables jointly affect the outcome. In such settings, piecewise constant estimation provides an intuitive approach by elucidating interactions and correlation patterns in addition to main effects. As well as generating 'and' statements similar to previously described methods, partDSA explores and chooses the best among all possible 'or' statements. The immediate benefit of partDSA is the ability to build a parsimonious model with 'and' and 'or' conjunctions. Currently, partDSA is capable of handling categorical and continuous explanatory variables and outcomes. This vignette provides a guide for analysis with the partDSA package while the actual algorithm is introduced and thoroughly described in Molinaro, Lostritto, and van der Laan (2010).

• Type: Journal
• Year: 2010
• Where: BIOINFORMATICS
• Authors: Annette M. Molinaro, Karen Lostritto, Mark J. van der Laan
• Keywords: BIOINFORMATICS 2010, Continuous Explanatory Variables, PartDSA Package, Piecewise Constant Estimation