BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.pydata.org//pydataglobal2025//talk//BT7M3S
BEGIN:VEVENT
UID:pretalx-pydataglobal2025-BT7M3S@cfp.pydata.org
DTSTART:20251210T133000Z
DTEND:20251210T140000Z
DESCRIPTION:In many regulated industries—finance\, healthcare\, insurance
 —logistic regression remains the model of choice for its interpretabilit
 y and regulatory acceptability. Yet capturing non-linear effects and inter
 actions often requires variable binning\, and naive approaches (equal-widt
 h or quantile cuts) can either wash out signal or invite overfitting. In t
 his 30-minute session\, data scientists and risk analysts with a working k
 nowledge of logistic regression and Python will learn to:\n\n-Diagnose the
  weaknesses of basic binning strategies.\n-Select and apply optimal-binnin
 g algorithms for different use cases.\n-Assess bin stability and guard aga
 inst model overfit.\n\nAll code\, data samples\, and a notebook will be av
 ailable on GitHub.
DTSTAMP:20260518T190809Z
LOCATION:Machine Learning & AI
SUMMARY:Optimal Variable Binning in Logistic Regression - Charaf ZGUIOUAR
URL:https://cfp.pydata.org/pydataglobal2025/talk/BT7M3S/
END:VEVENT
END:VCALENDAR