tidymodels for time-to-event data
summary
Demonstrates adding native time-to-event support to tidymodels by converting churn data to a Surv object, fitting proportional hazards or parametric/tree engines via parsnip, and wiring recipes + workflows + tune for resampling. Predicts both numeric survival time (median from the 50% threshold) and monthly probability of survival over 24 months per customer, keeps censored records instead of dropping them, and returns tibbles of per-person curves ready for plotting or threshold-based actions.
AI-written summary · about
chapters (5)
related talks
18mCensored - Survival Analysis in Tidymodels- 18mCensored - Survival Analysis in Tidymodels
18mtidymodels: Adventures in Rewriting a Modeling Pipeline
35mUsing Survival Analysis to understand customer retention
01h 24mBayesianism and Survival Analysis
31mHow to Validate you Client Churn Model
17mPrecision Medicine for All: Using Tidymodels to Validate PRS in Brazil
57mGood Practices for Applied Machine Learning