Tutorials

CogSci 2017: London

July 26 РJuly 29th, 2017

blank

Tutorials


Tutorials allow participants to gain new insights, knowledge, and skills from a range of topics in the field of cognitive science.
Many incorporate hands-on activities. The tutorials will be held the day before the main conference, on July 26, 2017. The location will be the same as the main conference.
 

Dynamic Field Theory: Conceptual Foundations and Applications in Cognitive and Developmental Science

Full day (9am – 4pm)

John Spencer, Vanessa Simmering, Sebastian Schneegans

Dynamical Systems thinking has been influential in the way psychologists, cognitive scientists, and neuroscientists think about sensori-motor behavior and its development. The initial emphasis on motor behavior was expanded when the concept of dynamic activation fields provided access to embodied cognition. Dynamical Field Theory (DFT) offers a framework for thinking about representation-in-the-moment that is firmly grounded in both Dynamical Systems thinking and neurophysiology. Dynamic Neural Fields are formalizations of how neural populations represent the continuous dimensions that characterize perceptual features, movements, and cognitive decisions. Neural fields evolve dynamically under the influence of inputs as well as strong neuronal interaction, generating elementary forms of cognition through dynamical instabilities. The concepts of DFT establish links between brain and behavior, helping to define experimental paradigms in which behavioral signatures of specific neural mechanisms can be observed. These paradigms can be modeled with Dynamic Neural Fields, deriving testable predictions and providing quantitative accounts of behavior.

 

Recent Advances in Deep Learning

Half day (Morning, 9am – 12pm)

Matthew Botvinick, Peter Battaglia

The present tutorial brings together five front-line researchers in AI, each with dual credentials in neuroscience and/or cognitive science, to provide an accessible overview and update on the most important recent developments in deep learning and deep RL. The tutorial will be aimed at a broad audience, ranging from graduate students to senior investigators, and spanning specialties from cognitive and developmental psychology to psychiatry, human factors research, and systems neuroscience. The focus will be on fundamental concepts and principles, and a central goal will be to maximize accessibility, in line with the tutorial format.

 

Methods for Reconstructing Causal Networks from Observed Time-Series: Granger-Causality, Transfer Entropy, and Convergent Cross-Mapping

Half day (Afternoon, 1-4pm)

Fermin Moscoso del Prado Martin

A major question that arises in many areas of Cognitive Science is the need to distinguish true causal connections between two variables from merely correlational ones. The most common way to address this issue is the design of well-controlled experiments. However, in many situations, however, it is extremely difficult or even outright impossible to perform such experiments. Researchers are then forced to rely on correlational data in order to make causal inferences. As is well-known this is often very problematic. From the correlations alone it is difficult to determine the direction of the causal arrow linking two variables. Worse even, the lack of controls of observational data entail that correlations found between two variables need not reflect any causal connection between them. It can very well be that some third variable for which we are either unaware of its importance, or just not able to measure it, is actually driving both our target variables, and in doing so giving rise to the mirage of a direct relationship between them.