Seminar on Neuroinformatics 5: Emotion in Machines and Humans
3
Algorithms and machine learning
Advanced studies
The seminar considers the phenomenon of emotions from a computational viewpoint.
| Year | Semester | Date | Period | Language | In charge |
|---|---|---|---|---|---|
| 2015 | spring | 12.01-27.04. | 3-4 | English | Aapo Hyvärinen |
Lectures
| Time | Room | Lecturer | Date |
|---|---|---|---|
| Mon 14-16 | B222 | Aapo Hyvärinen | 12.01.2015-23.02.2015 |
| Mon 14-16 | B222 | Aapo Hyvärinen | 09.03.2015-27.04.2015 |
Information for international students
The seminar will be in English.
General
The seminar considers the phenomenon of emotions from a computational viewpoint. It is highly multidisciplinary, and intended for MSc and PhD students in both computer science and neuroscience/psychology/cogsci. The questions considered include: What is an emotion, is it some kind of computation? Can computers have emotions? Can we understand human emotions by computer modelling? We go through a number of original scientific articles on these topics, the seminar participants giving talks on them.
Schedule:
(11 sessions + introduction, 2 talks per session)
12 Jan Introduction, allocation of papers and scheduling
19 Jan No meeting, no seminar (Preparation time for talks by yourself)
Basic psychological background:
26 Jan Joseph: An Argument for Basic Emotions. Mika: Core Affect and the Psychological Construction of Emotion.
Cognitive-computational theories:
02 Feb Luo: Motivational and emotional controls of cognition. Kaj: On the rationality of emotions: or, When are emotions rational?
09 Feb Munira: Theories of emotion causation: A review. Oskar: Computational models of emotion.
Artificial Intelligence models:
16 Feb Erkki: What does it mean for a computer to "have" emotions? Mikko: A domain-independent framework for modeling emotion.
23 Feb Yao: (paper changed) WASABI for affect simulation in human-computer interaction. Joel: Emergent dynamics of joy, distress, hope and fear in reinforcement learning agents
02 Mar Semester break
Evolutionary approaches:
09 Mar Quan: Evolutionary Psychology and the Emotions. Marta: What good are positive emotions?
Affective HCI:
16 Mar Luo: Computers that recognise and respond to user emotion. Oswald: Affect Detection: An Interdisciplinary Review.
23 Mar Janne: How emotion is made and measured and How to behave as Alice in Wonderland - about boredom and curiosity
Rupsha: Bodily maps of emotions.
Robotics:
30 Mar Vladimir: The uncanny advantage of using androids in cognitive ... research.
06 Apr: Easter
13 Apr: No seminar (lecturer travelling)
20 Apr Nicola: Why robots? A survey on ... social robots in the therapy of children with autism. Joosa: Robots that have emotions.
Gaming:
27 Apr Lauri: GAMYGDALA: an Emotion Engine for Games. Anna: Phasic emotional reactions to video game events:...
(Overflow papers: Flame -- fuzzy logic adaptive model of emotions. Automatic prediction of frustration. Affective Social Robots. Emotions are emergent processes. The following need some neuroscience background: The Biology of Fear. Emotions: from brain to robot )
Completing the course
To get the credit points, you are supposed to
- Participate in all the sessions of the seminar. One absence is OK, more than one if you have a very good reason.
- Give a talk based on given material, typically a scientific article. Who gets what will be decided in the first seminar session.
- Write, in the end, a short report on what you learned at the seminar (something like 10 lines per session, i.e. five lines per presentation). Deadline for report: 29th May.
- There is no overlap with the previous neuroinformatics seminars, so you can do this seminar every year.
Instructions for giving the talk:
Talks are 30 minutes long, and after the talk there is 15 minutes for discussion. Prepare slides using PowerPoint, OpenOffice Impress, Keynote (Mac) or with Latex using a similar document style (seminar, beamer). 15-20 slides is usually a reasonable number for a 30 minutes talk, although it depends a lot on the slides.The important point is not to have too much text on the slides, mainly figures and short lines with bullet points. An example of the style (not from a master's seminar) can be found in this slide show.
Registration:
Registration is closed, the seminar is full.
Literature and material
The articles considered in the seminar are given in the schedule above. (They were originally gathered on this page but it is just an archive with some short comments now).