Home Forums Kamanja Forums Data Science & Models R "caret" package – for streamlining creation of predictive models

This topic contains 0 replies, has 1 voice, and was last updated by  Greg Makowski 1 year, 9 months ago.

  • Author
  • #19834 Reply

    Greg Makowski

    For a best practice in creating predictive models in R, look at the caret package website.  Caret is short for Classification And REgression Training).  Note, this is not intended to cover all of data mining (such as clustering, association rules, optimization, NLP), just forecasting or prediction.

    Caret supports over 220+ algorithms (many packages support multiple algorithms).

    The standard R style documentation for caret currently covers 206 pages.

    The support is to provide a unified, consistent interface over all the algorithms.  For example, the model formula between dependent variable and independent variables is mostly consistently followed, but not always.  Caret covers

    * data splitting

    * pre-processing

    * feature selection

    * model tuning using resampling

    * variable importance estimation

    See the related book, Applied Predictive Modeling, at Amazon (4.8 / 5 stars from 51 reviews).

    • This topic was modified 1 year, 9 months ago by  Greg Makowski.
Reply To: R "caret" package – for streamlining creation of predictive models
Your information: