Research
Current research efforts as a Ph.D. student at the University of Utah include extending on high-performance scientific computing methods for simulating black hole mergers. More information about this research will be coming as papers are written and submitted.
Past Research
Master's Research
As a master's student at Brigham Young University, I studied underwater acoustics. In particular, I researched applications of maching learning for seabed type and source parameter predictions. Two different sound sources were used in my research: explosive charges and surface ships. A mixture of simulated samples and measured samples were used in both cases and I developed tools and programs to create the simulated data.
The explosive charge signals research was done in the time domain and were the subject of our initial experiments. We looked at using simple neural networks to make predictions on the plain time-domain waveforms and had some decent success. Greater success was found when using deep learning methods like convolutional neural networks even in the time domain. The work continued in the time domain and other studies focused on investigating how different simulated dataset parameters could improve predictions on measured data.
The surface ship research was done in the time-frequency domain via spectrograms. These spectrograms were used directly with convolutional neural networks to predict the type of sediment on the ocean floor as well as ship speed and the closest point of approach. The main findings were that creating sufficiently large synthetic datasets representative of the sea regions and source types of interest could be used to train convolutional neural networks to accurately predict seabed type and source parameters on measured signals even when a small amount of noise was present in the measured data.
Though my contributions to this field of research ended with my graduation, research on these problems continues with the research group at BYU.
Bachelor's Research
As a bachelor's student at Brigham Young University, I worked on applying acoustic beamforming methods to jet noise to analyze source properties. This research was fueled by the desire to understand the types of noise a tactical aircraft would produce to address concerns for personnel noise exposure and community annoyance and disturbance. Upon applying acoustic beamforming to the 71 element array of microphones as well as smaller subarrays, spatial overlap of competing noise sources was found and analyzed.