pull-bootstraps-generating-bootstrap-prediction.md

talk_id	talk_slug	talk_type	talk_tags	session_slug	sched_url	talk_title	talk_title_short	talk_materials_url	speakers
22033	pull-bootstraps-generating-bootstrap-prediction	regular	modeling tool	machine-learning	https://rstudioconf2022.sched.com/event/11iZW	Introducing workboots: Generate prediction intervals from tidymodel workflows	Introducing workboots	https://github.com/markjrieke/rstudio-conf-2022	name affiliation url username photo bio Mark Rieke Memorial Hermann Health System webpage twitter github linkedin affiliation https://www.thedatadiary.net/about/ https://twitter.com/markjrieke https://github.com/markjrieke https://www.linkedin.com/in/mark-j-rieke-ab4b0ab4/ https://www.memorialhermann.org/ mark_rieke /assets/img/2022Conf/_talks/22033_mark-rieke.png I am a senior consumer experience (CX) analyst at Memorial Hermann Health System where I use R and tidymodels to provide actionable insights from patient satisfaction survey data. I love making beautiful charts, working on home improvement projects, and playing jazzy piano. I live in Houston, TX, with my fiancé and two obnoxious yet lovable pets.

Sometimes, we want a model that generates a range of possible outcomes around each prediction. Other times, we just care about point predictions and may opt to use a fancy model like XGBoost. But what if we want the best of both worlds: getting a range of predictions while still using a fancy model? That’s where bootstrapping comes to the rescue! By using bootstrap resampling, we can create many models that produce a prediction distribution – regardless of the model type! In this talk, I’ll give an overview of bootstrap resampling for prediction, the pros/cons of this method, and how to implement it as a part of a tidymodel workflow with the workboots package.