Covers: theory of Diverse Ensembles Improve Calibration
0
Questions this item addresses:
  • How can we improve calibration of each ensemble member?
How to use this item?

The paper is fairly short, at 6 pages - I would recommend reading it through.

Author(s) / creator(s) / reference(s)
Asa Cooper Stickland, Iain Murray
0 comment
Recipe
publicShareStar

Calibrating Ensemble Members

Collaborators
Objectives
Learn about developments in ensemble member calibration
Potential Use Cases
More accurately assign probabilities to deep learning event outcomes.
Who is this for ?
ADVANCEDIf you're looking to ensure the confidence of your model's probability distributions.
Click on each of the following annotated items to see details.
ARTICLE 1. Ensemble Learning
  • What is ensemble learning?
10 minutes
PAPER 2. Diverse Ensembles Improve Calibration
  • How can we improve calibration of each ensemble member?
10 minutes
PAPER 3. Should Ensemble Members Be Calibrated?
  • How can we evaluate calibrated ensemble members?
10 minutes

Concepts Covered

0 comment