Probabilistic Calibration Illustrated by Probability Calibration Trees
summary
Presentation explains probability calibration using Probability Calibration Trees: it addresses the problem of making classifier probability outputs better calibrated. The talk describes the Probability Calibration Trees technique (a decision-tree-based method) and how it adjusts predicted probabilities to improve their reliability.
AI-written summary · about
chapters (9)
- 0:00Welcome
- 1:08Model calibration intro
- 7:10Calibration code demo
- 10:31Evaluating calibration
- 19:28Probability Calibration Trees
- 25:05Calibration methods
- 26:49Temperature scaling
- 29:59Why calibration matters
- 30:19Q&A
related talks
44mHandling predictive uncertainty in Machine Learning
01h 21mMulticalibration: Towards Fair Decision Making
35mUncertainty Unleashed
33mQuantifying uncertainty in machine learning models
36mQuantifying uncertainty in Machine Learning predictions
47mA Multi-Group Perspective on Fairness
31mUncertainty quantification: How much can you trust your machine learning model?
12mA Complexity-Theoretic Perspective on Fairness