Learning to Learn: What Machine Learning Can Tell Us About the Brain

Presented at the 2016 FAU Undergraduate Research Symposium

The vibrant and interdisciplinary field of machine learning has increasingly impacted many different fields of scientific study in the last 5-10 years. Current tools, such as neural networks, locally-competitive algorithms, and sparse coding have made it possible to gain insight into how the human brain analyzes and sorts the vast amounts of data it receives, as well as create machines that are capable of performing similarly, oftentimes better, than the human brain. In this simple demonstration inspired by the work of psychologist B.F. Skinner, we provide an illustration of the mechanism and power of machine learning by showing how, when implemented in an inexpensive rover with minimal hard-coding, they allow us to use operant conditioning to train the rover to perform certain behaviors in the presence of specific stimuli.



Compressed Sensing With a Dynamical Neural Network

Presented at FAU Graduate Research Day – March 2017


Compressed Sensing – Hyperspectral Satellite Imagery

Deep Learning Bird Recognition in Photographs

Presented at NCUR 2017