Abstract: R has an excellent framework for specifying models using formulas. While [elegant and useful](https://www.rstudio.com/rviews/2017/02/01/the-r-formula-method-the-good-parts/), it was designed in a time when models had small numbers of terms and complex preprocessing of data was not commonplace. As such, it has some [limitations](https://www.rstudio.com/rviews/2017/03/01/the-r-formula-method-the-bad-parts/). In this talk, a new package called `recipes` is shown where the specification of model terms and preprocessing steps can be enumerated sequentially. The recipe can be estimated and applied to any dataset. Current options include simple transformations (log, Box-Cox, interactions, dummy variables, ...), signal extraction (PCA, ICA, MDS), basis functions (splines, polynomials), imputation methods, and others.
Bio: Max Kuhn works at RStudio developing software for data analysis and modeling. He previously worked in pharmaceutical and molecular diagnostic research for more than 18 years. Max’s interests are in predictive modeling and machine learning and is the author of six R packages, including the [caret package](http://topepo.github.io/caret/). He and Kjell Johnson published the bestselling book [Applied Predictive Modeling](http://appliedpredictivemodeling.com) in 2013. Max holds a B.S. in Mathematics and a Ph.D. in Biostatistics.
Software Engineer at RStudio