As a scientist and an engineer, I am fascinated with information: what it is,
how it is stored, how it is accessed, and how it leads to decisions. My doctoral
research focused on how patterns in disk accesses allow a system to predict what
data is commonly accessed together along with how this knowledge of momentary
grouping can be used to make systems more available and power efficient.
At Emory, I founded the SimBioSys
Lab where I, along with 2 other faculty members and our students, explore
the intersections of systems, biology, and large simulated environments. My
particular interest is the dual of the storage problem. Whereas computer
scientists have defined how to arrange storage to meet specific metrics such as
fault tolerance and access speed, in neuroscience the metrics are observable
but the system unknown. We are working to model information in the brain as a
storage problem to better learn how we collect and interpret signals from our
world, working towards a robust fault tolerance model for the brain.
Other research interests in our lab include machine learning applications in neurobiology (particularly deep networks and topological data mining), archival storage, power management, privacy, and spatiotemporal modeling of storage workloads.
My Erdős number is 3.
Interdisciplinary Work at Emory,
Archival Storage (The Register),
about my work at Salk
Undergraduate Recommendation Policy: (adapted from Amy Weldon)
I don't write a reference for a student unless I can write a very
positive and specific one. Therefore, your job as a college student is to
become the kind of student professors can rave about in recommendations:
hardworking, collegial, and intellectually inquisitive and honest. Consider
maintaining relationships over time with professors, so that they know you well
enough to write for you. Many juniors and seniors tell me they wish they had
thought about this during their first year.
This word cloud represents popular words in my published papers as of 2013. More details about my work are on my Research Page.