Boosting

Posted in Science on September 06, 2008


Boosting

Lecture slides:

  • An Introducting to Boosting
  • Contents of this Course
  • Sources of Information
  • Learning - Problem Formulation
  • Example: natural Apples vs. plastic Apples
  • Simple Hypotheses: Cuts on Coordinate Axes
  • PAC Learning
  • Levels of Generality
  • Weak and Strong Learning
  • 2. Basic Issues in Boosting
  • Simple Hypotheses: Cuts on Coordinate Axes.
  • AdaBoost (Freund&Schapire 1996)
  • Boosting: 1st Iteration (simple hypothesis)
  • Boosting:Recompute weighting
  • Boosting: 2nd Iteration
  • Boosting: 2nd Hypothesis
  • Boosting: Recompute weighting.
  • Boosting: 3rd Hypothesis
  • Boosting: 4th Hypothesis
  • Combination of Hypotheses
  • Decision
  • Ensemble learning for Classification
  • Early Algorithms
  • AdaBoost algorithm
  • Desirable Properties of the Weak Learner
  • Weak Learners used with Boosting
  • Hypothesis weight
  • Reweighting
  • What is going to come in the next Session?
  • Some Highlights

Author: Gunnar Rätsch, Max Planck Institute

Watch Video

Tags: Science, Lectures, Computer Science, Machine Learning, VideoLectures.Net, Boosting