Bamberg University

 k-Nearest Neighbor  Locally weighted regression  Radial basis functions  Case-based reasoning  Lazy and eager learning

 Sample error, true error  Condence intervals for observed hypothesis error  Estimators  Binomial distribution, Normal distribution, Central Limit Theorem  Paired t tests  Comparing learning methods

 Control learning  Control policies that choose optimal actions  Q learning  Convergence

 Why combine inductive and analytical learning?  KBANN: Prior knowledge to initialize the hypothesis  TangetProp, EBNN: Prior knowledge alters search ob jective  FOCL: Prior knowledge alters search operators

• Two formulations for learning: Inductive and Analytical • Perfect domain theories and Prolog-EBG

 Learning from examples  General-to-specic ordering over hypotheses  Version spaces and candidate elimination algorithm  Picking new examples  The need for inductive bias

 Sequential covering algorithms  FOIL  Induction as inverse of deduction  Inductive Logic Programming

 Evolutionary computation  Prototypical GA  An example: GABIL  Genetic Programming  Individual learning and population evolution

 Computational learning theory  Setting 1: learner poses queries to teacher  Setting 2: teacher chooses examples  Setting 3: randomly generated instances, labeled by teacher  Probably approximately correct (PAC) learning  Vapnik-Chervonenkis Dimension  Mistake bounds

 Bayes Theorem  MAP, ML hypotheses  MAP learners  Minimum description length principle  Bayes optimal classier  Naive Bayes learner  Example: Learning over text data  Bayesian belief networks  Expectation Maximization algorithm