Name of the study programme component: - ESAT

stepweedheightsAI and Robotics

Oct 15, 2013 (3 years and 9 months ago)

73 views

Name of the study programme component:

Machine Learning and Inductive Inference (H0D85B)


Teacher responsible:

Hendrik Blockeel


Education form:



Contact hours



Budgeted study time
1

-

lecture



19,5

hours



.... hours

-

seminar



.... hours



.... hour
s

-

exercise



13

hours



.... hours

-

practical



….

hours



.... hours

-

other




….

hours



.... hours

-





.... hours



.... hours

-





.... hours



.... hours

Total




32,5

hours



.... hours


Examination method:


O
ral
exam with written preparation
.

The exam consists of questions about the theory as well as
some exercises.

Brief description of the objectives, content of the study programme component and interpretation
of education forms:


Aims:

This course aims at familiarising the students with th
e domain of machine learning, which
concerns techniques to build software that can learn how to perform a certain task (or improve its
performance on it) by studying examples of how it has been accomplished previously, and in a
broader sense the discovery
of knowledge from observations (inductive inference).

The course introduces some general principles of machine learning and inductive inference, and
gives an overview of the existing approaches and al
gorithms.


Content:

-

introduction to machine learning,
connections with other domains

-

general principles of learning

-

concept learning and version spaces

-

evaluation of learning algorithms

-

computational learning theory

-

representational aspects

-

specific approaches to learning

-

induction of decision t
rees

-

induction of rule sets and association rules

-

inductive logic programming

-

neural nets

-

Bayesian approaches

-

instance based learning

-

genetic algorithms

-

reinforcement learning

-

support vector machines

-

combi
ning models (bagging, boosting,
..
.)


Literature consulted:

-

T. Mitchell: Machine Learning. Published by McGraw
-
Hill.






1

The results of a study measurement may be included in the table if the study programme has them.