Date of Submission

Spring 2019

Academic Programs and Concentrations

Psychology; Psychology

Project Advisor 1

Justin C. Hulbert

Abstract/Artist's Statement

Learning is a ubiquitous process that transforms novel information and events into stored memory representations that can later be accessed. As a learner acquires new information, any feature of a memory that is shared with other memories may produce some level of retrieval- competition, making accurate recall more difficult. One of the most effective ways to reduce this competition and create distinct representations for potentially confusable memories is to practice retrieving all of the information through self-testing with feedback. As a person tests themself, competition between easily-confusable memories (e.g. memories that share similar visual or semantic features) decreases and memory representations for unique items are made more distinct. Using a portable, consumer-grade electroencephalography (EEG) device, I attempted to harness competition levels in the brain by training a machine learning classifier to predict long- term retention of novel associations. Specifically, I compare the accuracy of two logistic regression classifiers: one trained using existing category-word pairings (as has been done previously in the literature), and one trained using new episodic image-name associations developed to more closely model memory competition. I predicted that the newly developed classifier would be able to more accurately predict long-term retention. Further refinements to the predictive model and its applications are discussed.

Open Access Agreement

Open Access

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS