# Probabilistic reasoning

Course code:INFOPROB
Credits:7.5 ECTS (=5.25 old credit points)
Period:periode 1 (week 36 t/m 46, dwz 2-9-2004 t/m 12-11-2004; herkansing week 52)
Timeslot:C
Participants:up till now 26 subscriptions
Schedule:Dit is een oud rooster!
formgrouptimeweekroomteacher
college   ma 13-1537-45 BBL-505 Silja Renooij

do 09-1136-45 BBL-505
werkcollege   di 15-1737-45 BBL-505 Janneke Bolt

Contents:Human experts can make judgments and decisions based on uncertain, and often even conflicting, information. A knowledge-based system that is required to perform at least at a similar level of expertise, should be able to cope with this type of information. For this reason, formalisms for representing uncertainty and algorithms for manipulating uncertain information are important research subjects within the field of Artificial Intelligence. Probability theory is one of the oldest theories dealing with the concept of uncertainty; it is therefore no surprise that the applicability of this mathematical theory as a model for reasoning under uncertainty plays an important role. In this course, we will consider probabilistic models for manipulating uncertain information in knowledge-based systems. More specifically, we will consider the theory underlying the framework of probabilistic networks, and discuss the construction of such networks for real-life applications.
Literature:Syllabus 'Probabilistic Reasoning' and transparencies (both online).
Course form:Lectures (twice a week); Excercise class (once a week).
Exam form:One written exam (closed book!).
Minimum effort to qualify for 2nd chance exam:Om aan de aanvullende toets te mogen meedoen is ontbreken van ten hoogte 1 toetsactiviteit toegestaan.
Description:In this course, we will consider the theory and applicability of probabilistic networks. The course roughly consists of three parts. As an introduction to probabilistic networks, the first part of the course deals with independence relations and their graphical represenation by means of undirected and directed graphs. The second part introduces the probabilistic network as a compact representation of a probability distribution on a set of statistical variables; in addition, the Pearl algorithm for computing probabilities from a probabilistic network is discussed. The algorithm allows for calculating the probability of any value of an arbitrary variable in the network, with or without incorporating observations for one or more variables. The third part of this course concerns the construction of probabilistic networks for real-life applications. Topics covered include both automated construction of networks from data, and handcrafting the network with the help of domain experts.
wijzigen?