Ben Baker's

Homepage

My research typically pertains to computational seismology. With regards to University of Utah Seismograph Stations this means developing the business logic to deliver data products of higher-quality more consistently and more expediently. This is achieved by packaging theoretical and practical seismological principles into automatic decision-making frameworks.

Most projects with which I am involved are available at the UUSS github page. In addition to maintaining our production AQMS system, I primarily work on operationalizing machine-learning algorithms. This requires first creating a messaging system (UMPS) that then enables machine-learning algorithms to function in a real-time setting (URTS).

Detecting small earthquakes is important in the context of generating complete earthquake catalogs. An effective strategy for finding tiny temblors missed during routine processing involves scanning through continuous data and looking for waveform snippets that look like known earthquake signals. This process is known as matched filtering and is implemented in MFLib available here.

To enhance feature extraction for machine learning algorithms and expedite routine seismic processing I am perpetually extending a library for real-time processing and post-processing of seismic signals. RTSeis is available here.

GEO 5920/6920: Statistical Methods Applied to Earthquake Seismology. This is a machine learning class which is specialized to regional network seismology.

Periodically UUSS makes available undergraduate research positions. If you are mathematically inclined and interested in computing then please email me to learn about current undergraduate student research opportunities.

My CV is available here.