CGSC 4605
Hyperdimensional Cognitive Models
Modelling cognition using artificial intelligence techniques such as reinforcement learning, vector-symbolic models, neural networks, and/or machine learning.
2024
Topics
- Linear Algebra
- Holographic Vectors
- Rule Induction Rasmussen & Eliasmith (2011) A Neural Model of Rule
- Neural Engineering Eliasmith, Tang, & DeWolf (2013) How to build a brain
- Memory & Learning Franklin & Mewhort (2015) Memory as a hologram
- Memory Tesseract Levy & Gayler (2009) “Lateral Inhibition”
- Semantic Space Landauer & Dumais (1997) A Solution to Plato’s Problem
- Last day for academic withdrawal
- Declarative Memory Kelly et al. (2020) Holographic Declarative Memory
- Catastropic Mannering & Jones (2021) Catastrophic Interference in
- Transformers Vaswani et al. (2017) Attention is all you need
Prerequisites & course information
Prerequisite(s): third-year standing, ( CGSC 1005 or COMP 1005 ), CGSC 2001, and CGSC 3601 . Also offered at the graduate level, with different requirements, as CGSC 5605 , for which additional credit is precluded. Seminar three hours per week, tutorial one and a half hours per week.