Despite decades of research into the nature of memory, many basic
questions about the mechanisms of learning, information storage and
retrieval in the human brain remain unanswered. In the proposed
year-long RIBMS project, a team of two undergraduate students will
work with an interdisciplinary team at the Memory and Cognition Center
at Case Western Reserve University to study fundamental mechanisms of
memory based on analysis of controlled word association and recall
tasks. Specifically, the students will develop computational models
and analyze experimental data to explore the concept of semantic space
from both mathematical and biological perspectives.
Through the Memory and Cognition Center databases, students will
analyze data on both category fluency and verbal recall in individuals
with varying degrees of cognitive impairment. The standard approach to
such datasets is to use logistic regression models to compare groups
while correcting for demographic factors such as age and
education. However, such an approach is limited insofar as it does not
consider the path through the interconnected network that can
represent semantic space. A graph theoretical approach would utilize
network metrics using data from multiple individuals who had completed
the same task. Implications for network theory and biological reality
will be explored and additional experiments may be considered.
Directed reading in network and information theory (e.g. Albert and
Barabasi, Shannon) will provide background for understanding the
relevance of the approach to the data.
Powerpoint slides
from project introduction presentation
|
