(→Syllabus) |
(→Syllabus) |
||
(12 intermediate revisions by the same user not shown) | |||
Riadok 30: | Riadok 30: | ||
|Tuesday | |Tuesday | ||
|09:50 - 11:20 | |09:50 - 11:20 | ||
− | |I- | + | |I-23 |
|[[Igor Farkas|Igor Farkaš]] | |[[Igor Farkas|Igor Farkaš]] | ||
|- | |- | ||
Riadok 36: | Riadok 36: | ||
|Tuesday | |Tuesday | ||
|11:40 - 13:05 | |11:40 - 13:05 | ||
− | |I- | + | |I-23 |
|students | |students | ||
|} | |} | ||
Riadok 57: | Riadok 57: | ||
|- | |- | ||
|(2) <br>26.09. | |(2) <br>26.09. | ||
− | |Math concepts for DL in a nutshell | + | |Math concepts for DL in a nutshell <b>(Števo) </b> |
| | | | ||
♦ Zhang A. et al. (2020). [https://d2l.ai/ Dive into Deep Learning]. An interactive deep learning book with code, math, and discussions, based on the NumPy interface. Appendix 22: Mathematics for Deep Learning.<br> | ♦ Zhang A. et al. (2020). [https://d2l.ai/ Dive into Deep Learning]. An interactive deep learning book with code, math, and discussions, based on the NumPy interface. Appendix 22: Mathematics for Deep Learning.<br> | ||
Riadok 65: | Riadok 65: | ||
|Current hype in deep learning | |Current hype in deep learning | ||
| | | | ||
− | ♦Perconti P., Plebe A. (2020) [https://doi.org/10.1016/j.cognition.2020.104365 Deep learning and cognitive science]. <i>Cognition</i>, 203 <b>[ | + | ♦Perconti P., Plebe A. (2020) [https://doi.org/10.1016/j.cognition.2020.104365 Deep learning and cognitive science]. <i>Cognition</i>, 203 <!--b>[Tina & Maša]</b--> |
<!--=============================================================================================--> | <!--=============================================================================================--> | ||
|- | |- | ||
|(4) <br>10.10. | |(4) <br>10.10. | ||
− | |Deep learning in computer vision | + | |Deep learning in computer vision. Example of simple image classification tasks in Python (Števo) |
<!--a href="Slides/common-coding.4x.pdf">slides</a--> | <!--a href="Slides/common-coding.4x.pdf">slides</a--> | ||
| | | | ||
− | ♦ Lindsay G.W. (2021) [https://doi.org/10.1162/jocn_a_01544 Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future]. <i>Journal of Cognitive Neuroscience</i>, 33(10): 2017–2031 <b>[ | + | ♦ Lindsay G.W. (2021) [https://doi.org/10.1162/jocn_a_01544 Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future]. <i>Journal of Cognitive Neuroscience</i>, 33(10): 2017–2031 <!--b>[Tiziana]</b--> |
<br> | <br> | ||
<!--=============================================================================================--> | <!--=============================================================================================--> | ||
|- | |- | ||
|(5) <br>17.10. | |(5) <br>17.10. | ||
− | |Deep learning in natural language processing | + | |Deep learning in natural language processing. Examples in Python <b>(Marek)</b> |
<!--a href="Slides/lang-action.4x.pdf">slides</a--> | <!--a href="Slides/lang-action.4x.pdf">slides</a--> | ||
| | | | ||
− | ♦ DeepLearning.AI (2023). [https://www.deeplearning.ai/resources/natural-language-processing/ A Complete Guide to Natural Language Processing]. < | + | ♦ DeepLearning.AI (2023). [https://www.deeplearning.ai/resources/natural-language-processing/ A Complete Guide to Natural Language Processing]. <!--b>[Jaš]</b--> |
<!--=============================================================================================--> | <!--=============================================================================================--> | ||
|- | |- | ||
Riadok 86: | Riadok 86: | ||
|Deep learning for robotics | |Deep learning for robotics | ||
<!--a href="Slides/barsalou.lass.4x.pdf">slides</a--> | <!--a href="Slides/barsalou.lass.4x.pdf">slides</a--> | ||
− | |Sünderhauf N, Brock O, Scheirer W, et al. (2018) [https://doi.org/10.1177/0278364918770733 The limits and potentials of deep learning for robotics]. <i>The International Journal of Robotics Research</i>, 37(4-5), 405-420 | + | |Sünderhauf N, Brock O, Scheirer W, et al. (2018) [https://doi.org/10.1177/0278364918770733 The limits and potentials of deep learning for robotics]. <i>The International Journal of Robotics Research</i>, 37(4-5), 405-420 <!--b>[Jasper & Larissa]</b--> |
<!--=============================================================================================--> | <!--=============================================================================================--> | ||
|- | |- | ||
|(7) <br>31.10. | |(7) <br>31.10. | ||
− | |Deep reinforcement learning | + | |Deep reinforcement learning. Example of a simple task in Python <b>(Iveta)</b> |
<!--a href="Slides/covariation.4x.pdf">slides</a--> | <!--a href="Slides/covariation.4x.pdf">slides</a--> | ||
| | | | ||
− | ♦ Arulkumaran K., Deisenroth M.P.,Brundage M., Bharath A.A. (2017) [https://arxiv.org/pdf/1708.05866.pdf Brief Survey of Deep Reinforcement Learning]. <i>IEEE Signal Processing Magazine</i>, 26-38 <b>[ | + | ♦ Arulkumaran K., Deisenroth M.P.,Brundage M., Bharath A.A. (2017) [https://arxiv.org/pdf/1708.05866.pdf Brief Survey of Deep Reinforcement Learning]. <i>IEEE Signal Processing Magazine</i>, 26-38 <!--b>[Barbara]</b> <br--> |
<!--============================================================================================= | <!--============================================================================================= | ||
|- | |- | ||
Riadok 106: | Riadok 106: | ||
<!--a href="Slides/louwerse.sihypo.4x.pdf">slides</a--> | <!--a href="Slides/louwerse.sihypo.4x.pdf">slides</a--> | ||
| | | | ||
− | ♦ Huang S. et al. (2023): [https://arxiv.org/pdf/2302.14045.pdf Language Is Not All You Need: Aligning Perception with Language Models]. arXiv. <b>[ | + | ♦ Huang S. et al. (2023): [https://arxiv.org/pdf/2302.14045.pdf Language Is Not All You Need: Aligning Perception with Language Models]. arXiv. <!--b>[Tim & Arina & Pia]</b--> |
<br> | <br> | ||
<!--=============================================================================================--> | <!--=============================================================================================--> | ||
Riadok 114: | Riadok 114: | ||
<!--a href="courses/GroundedCog/Slides/farkas-etal.icub.4x.pdf">slides</a--> | <!--a href="courses/GroundedCog/Slides/farkas-etal.icub.4x.pdf">slides</a--> | ||
| | | | ||
− | ♦ Zhu Y. et al. (2020): [https://doi.org/10.1016/j.eng.2020.01.011 Dark, Beyond Deep: A Paradigm Shift to Cognitive AI with Humanlike Common Sense]. <i>Engineering</i>, 6(3), pp. 310-345 <b>[ | + | ♦ Zhu Y. et al. (2020): [https://doi.org/10.1016/j.eng.2020.01.011 Dark, Beyond Deep: A Paradigm Shift to Cognitive AI with Humanlike Common Sense]. <i>Engineering</i>, 6(3), pp. 310-345 <!--b>[Klara & Grega & Rick]</b><br--> |
<!--=============================================================================================--> | <!--=============================================================================================--> | ||
|- | |- | ||
Riadok 121: | Riadok 121: | ||
| | | | ||
♦Garnelo, M. and Shanahan, M. (2019): [https://doi.org/10.1016/j.cobeha.2018.12.010 Reconciling deep learning with symbolic artificial intelligence: representing objects and relations]. <i>Current Opinion in Behavioral Sciences</i>, 29(17), 2352-1546 | ♦Garnelo, M. and Shanahan, M. (2019): [https://doi.org/10.1016/j.cobeha.2018.12.010 Reconciling deep learning with symbolic artificial intelligence: representing objects and relations]. <i>Current Opinion in Behavioral Sciences</i>, 29(17), 2352-1546 | ||
− | <b>[ | + | <!--b>[Omar + Ana]</b> <br--> |
<!--============================================================================================= | <!--============================================================================================= | ||
|- | |- | ||
− | |(11) <br>28. | + | |(11) <br>28.11. |
|Integrative summary | |Integrative summary | ||
| | | |
Aktuálna revízia z 09:37, 25. jún 2024
Deep Learning and Cognition – 2-IKV-239a/22
Obsah
The course introduces the field of a highly popular machine learning approach focused on deep learning in artificial neural networks. Aiming at master’s students with diverse backgrounds of bachelor’s studies (such as students of cognitive science), the course will guide them through different areas of DL, to highlight its successful applications. To better understand the mechanistic principles of DL models, the students will learn the basics about the underlying mathematical concepts of DL and will be shown a few simpler examples of functioning neural network models. Throughout the course, the discussions will also focus on virtues and vice of deep learning and its relation to human cognition.
The course is a part of Master Programme in Cognitive Science.
Course schedule
Type | Day | Time | Room | Lecturer |
---|---|---|---|---|
Lecture | Tuesday | 09:50 - 11:20 | I-23 | Igor Farkaš |
Student presentation | Tuesday | 11:40 - 13:05 | I-23 | students |
Syllabus
Date | Topic | References |
---|---|---|
(1) 19.09. |
Introduction to human intelligence and artificial intelligence |
♦ Korteling J.E. et al. (2021) Human- versus Artificial Intelligence. Frontiers in Artificial Intelligence, 4 |
(2) 26.09. |
Math concepts for DL in a nutshell (Števo) |
♦ Zhang A. et al. (2020). Dive into Deep Learning. An interactive deep learning book with code, math, and discussions, based on the NumPy interface. Appendix 22: Mathematics for Deep Learning. |
(3) 03.10. |
Current hype in deep learning |
♦Perconti P., Plebe A. (2020) Deep learning and cognitive science. Cognition, 203 |
(4) 10.10. |
Deep learning in computer vision. Example of simple image classification tasks in Python (Števo) |
♦ Lindsay G.W. (2021) Convolutional Neural Networks as a Model of the Visual System: Past, Present, and Future. Journal of Cognitive Neuroscience, 33(10): 2017–2031
|
(5) 17.10. |
Deep learning in natural language processing. Examples in Python (Marek) |
♦ DeepLearning.AI (2023). A Complete Guide to Natural Language Processing. |
(6) 24.10. |
Deep learning for robotics | Sünderhauf N, Brock O, Scheirer W, et al. (2018) The limits and potentials of deep learning for robotics. The International Journal of Robotics Research, 37(4-5), 405-420 |
(7) 31.10. |
Deep reinforcement learning. Example of a simple task in Python (Iveta) |
♦ Arulkumaran K., Deisenroth M.P.,Brundage M., Bharath A.A. (2017) Brief Survey of Deep Reinforcement Learning. IEEE Signal Processing Magazine, 26-38 |
(8) 07.11. |
Multimodal models |
♦ Huang S. et al. (2023): Language Is Not All You Need: Aligning Perception with Language Models. arXiv.
|
(9) 14.11. |
Going beyond deep learning |
♦ Zhu Y. et al. (2020): Dark, Beyond Deep: A Paradigm Shift to Cognitive AI with Humanlike Common Sense. Engineering, 6(3), pp. 310-345 |
(10) 21.11. |
Toward cognitive AI |
♦Garnelo, M. and Shanahan, M. (2019): Reconciling deep learning with symbolic artificial intelligence: representing objects and relations. Current Opinion in Behavioral Sciences, 29(17), 2352-1546 |
(11) 28.11. |
Group presentations | groups of 3-4 students, presenting a relevant topic |
Grading
- Weekly activity during the semester (50%). This includes weekly submitting inputs to the moderator and active participation during discussions.
- Paper presentation and discussion moderation (30%). You will present a selected topic (one of the papers in the syllabus), collect by email the inputs (one question or a discussion point) from other students in advance (until Sunday, 20:00). The inputs should will sent to the moderator, with the subject "author" (use the first author's surname). The moderator organizes the questions for discussion that he/she will moderate in the class.
- Final group presentation (20%). You will be organized in small groups (3-4 students) and will prepare a final presentation on the topic of your choice relevant for the course.
- Overall grading (in %): A > 90, B > 80, C > 70, D > 60, E > 50, else Fx.
Attendance
The class will be run in a classical (in-person) form. You are expected to regularly attend the class. In case of absence, inform the teacher about the reason. Max. two absences are ok (i.e. without losing points).