Probabilistic Reasoning

Class and subject schedule

Check this page regularly, for the overview of subjects may be updated during the course.

The table below lists:

the scheduled lectures and other types of session, together with their topics
the mandatory literature to study in addition to the slides
(Note: we generally advise you to read the literature prior to the lecture and review it directly afterwards. In cases where additional preparation is necessary, this is explicitly stated.)
the exercises to train (in lecture notes); you are strongly encouraged to do these right before or after class so you can ask questions about them at our next meeting, if needed.

The literature referred to can be found on the literature page, which also includes (a.o.):

the course slides (these are mandatory material);
🎦 videos associated with some of the lectures.

The scheduled times and locations for the lectures and exams can be found in MyTimetable. Since the lecturers cannot directly update MyTimetable, deviations from the official schedule will be indicated in the schedule below, which is always most up-to-date.

Week 1 (36)

Wed, Sept 3

Subjects: Introduction uncertain reasoning and PGMs

Literature: Lecture notes Ch. 1, Ch. 2 + slides Ch.1,2

Exercises (in Lecture Notes): 1.1, 1.2, 2.1 - 2.6 (to refresh probability theory background)

Assignments (on Brightspace): Already start working on the Practical assignments! (Can't find a partner? Try Brightspace's discussion forum.)

Fri, Sept 5

Preparation: make sure that you are fluent in the pre-requisite background reviewed in Ch. 2 (also partially and briefly covered in this video: 🎦 Ch2); the whole course builds upon this! Also: familiarize yourself with the notations used in this course.

Subjects: Independence relations

Literature: Lecture notes Ch. 2, Ch. 3: § 3.1, [optional: 3.2.1] + slides Ch.3

Exercises: 2.7 - 2.12, 3.1, [optional: 3.2, 3.3], 3.4

Week 2 (37)

Wed, Sept 10

Preparation: review the literature on independence relations, possibly with the help of this video: 🎦 Ch3.1

Subject: Graphical models

Literature: Lecture notes Ch. 3: § 3.2.2, 3.2.3 + slides Ch.3

Exercises: [optional: 3.5, 3.6], 3.7-3.14

Fri, Sept 12

Subject: Bayesian networks & introduction to inference

Literature: Lecture notes Ch. 4: § 4.1 + slides Ch.4

Exercises: 4.1, 4.2, 4.8

Week 3 (38)

Wed, Sept 17

Deadline Assignment A (directly after class)

Subject: Inference in singly connected graphs (SCGs) and trees

Literature: Lecture notes Ch. 4: § 4.2.1, § 4.2.2 + slides Ch.4

Exercises: 4.3 - 4.5

Directly after class: 🎦review proofs for computation rules. It is also really important to train yourself with the exercises!

Fri, Sept 19

Subject: Examples in SCGs; bridge to Bayesian approach

Literature: Lecture notes Ch. 4: § 4.2.1, § 4.2.2, § 4.2.4, [optional, but strongly advised: Notes on Probabilistic Models] + slides Ch.4, Bridge-to-PP

Exercises: 4.6, 4.7, 4.13

Week 4 (39)

Wed, Sept 24

Preparation: Start preparing Assignment C: we assume that you are able to set up all software on your favorite OS by yourself, or with help from your fellow students
Subject: Probabilistic programming, part I (Bayesian data analysis and continuous models)

Literature: Book BDA3 Ch1 up to and including § 1.7 + slides PP1

Fri, Sept 26

Subject: Probabilistic programming, part II (inference in continuous models)

Literature: Paper RR04 up to and including § 3.2 (skim the rest) + slides PP2

Week 5 (40)

Wed, Oct 1

Double slot, starting at 1pm! First a lecture, then a Stan practical

Preparation: really make sure to have Stan up and running on your computer before today's sessions to ensure that you can benefit from them for Assignment C!

See the Stan documentation

Getting started with RStan; for this you will need to install RStan

If you like to use Stan from Python, try CmdStanPy

🎦watch the following videos from the playlist of this Stan tutorial: 1 or 2 (depending on your choice for R or Python), and 3--7

Note that Stan supports every C++ compiler except for the Windows MSVC one (which doesn't adhere to conventions). Windows users therefore need to use a different C++ compiler such as mingw. Alternatively: use Linux or OSX.

Subjects: Probabilistic programming, part III (Workflow and practical session)

Literature: same as above + slides PP3

Fri, Oct 3

Subject: Inference in multiply connected graphs (MCGs)

Literature: Lecture notes Ch. 4: § 4.2.3, § 4.2.4 + slides Ch.4

Exercises: 4.9-4.11, 4.12, 4.14

Afterwards: really train yourself with the exercises!

[Optional: additional slides for those interested in Junction Tree propagation]

Week 6 (41)

Wed, Oct 8

Subject: Graph construction

Literature: Lecture notes Ch. 5: §5.1, §5.2.1 and § 5.2.2 + slides Ch.5

Exercises: 5.1-5.3, 5.4a

Fri, Oct 10

Deadline Assignment B (directly after class)

Subject: construction cntd

Literature: Lecture notes Ch. 5: § 5.2.2 and § 5.3.1 + slides Ch.5

Exercises: 5.4b, 5.7, 5.8

Week 7 (42)

Wed, Oct 15

Subject: Quantification; Noisy-or gate

Literature: Lecture notes Ch. 5: § 5.3.1, 5.3.2, 5.3.3 + slides Ch.5

Exercises: 5.4cd, 5.9, 5.11

Fri, Oct 17

Subject: Probability elicitation methods, Sensitivity analysis

Literature: Lecture notes Ch. 5: § 5.3.3 and § 5.3.4, Ch. 6: § 6.1.1 + slides Ch.5,6

Exercises: 5.5, 5.6, 5.10, 6.1a-d, 6.2ab, 6.7

Week 8 (43)

Wed, Oct 22

Double slot, starting at 1pm! First coaching for assignment C, then a lecture.

Preparation for coaching/ Q&A: prepare all your remaining questions about the assignment

Subject lecture: Sensitivity Analysis cntd

Literature: Lecture notes Ch. 6: 6.1.2, § 6.1.3 + slides Ch.6

Exercises: 6.1ef, 6.2cde, 6.3-6.6

Fri, Oct 24

Deadline Assignment C (provisional)

Subject: Evaluation; BNs as problem-solving architectures

Literature: Lecture notes Ch 6: § 6.2, 6.3 + slides Ch.6

Exercises: 6.8

Week 9 (44)

Wed, Oct 29

Preparation: prepare your final questions, this could be the last lecture!

Subject: Explanation; possibly wrap-up and Q & A

Literature: Lecture notes Ch 6: § 6.4; Ch. 7 + slides Ch.6,7

[Optional: Jordy van Leersum's short explanation demo]

Fri, Oct 31

EXTRA SLOT

Disclaimer: will only be used if we deviate from the above schedule. Since the exam is in less than a week, any subjects treated for the first time today are not part of the examination.

Week 10 (45)

Wed Nov 5 (check MyTimetable!)

Written test ('Tentamen')

January 2025

(check MyTimetable!)

Substitute test / retake ('Tentamen: aanvullende toets')
Substitute tests / retakes for the masters' courses will (probably) be scheduled in week 3, in the late afternoon/evening. Once scheduled, exams will appear in MyTimetable; this is also the source of information for the lecturers.)