Brian Goldman

acxa86@r.postjobfree.com

703-***-****

Technical

Expertise in the design and implementation of algorithms for:

● Machine learning , including regularized regression, principal component analysis, deep learning, natural language processing, data clustering, and ensemble methods.

● Optimization, including stochastic processes, linear and nonlinear programming, network flow, and discrete optimization heuristics.

● Linear algebra, including exact and fast approximate computation methods for matrix factorizations, inverses, polynomials, and products. Experienced applying these algorithms to problems such as resource allocation, user preference prediction, financial management, and social network analysis. Additional competencies include web scraping, visualization, distributed computing, database query (SQL and Pandas), applied probability, and discrete-event simulation.

Experience

The Data Incubator

● Data Scientist in Residence (2015)

--Launched The Data Incubator’s program in Kuala Lumpur, Malaysia; taught two course sessions, advised on client projects, evaluated course applicants and technical staff.

● Fellow (2015)

--Applied techniques from natural language processing and distributional statistics to search for evidence of corporate fraud in SEC filings (independent capstone).

--Developed and deployed Python solutions for projects in sentiment analysis, time series forecasting, celebrity influence modeling, and parallel processing. Education

College of William and Mary

● Master of Science in Computer Science (2015), Bachelor of Science cum laude (2012).

--Specialization in Computational Operations Research (MS).

--Mathematics major, Biology minor (BS).

--Graduate Studies and Research Recruitment Fellowship (2013-2015).

--National Merit Finalist (2008).

● Teaching Assistantship (2011, 2013-2015).

--Graduate Courses: Probability and Linear Programming. Upper Division: Elementary Analysis and Linear Algebra. Introductory: Calculus I and II. Research

● Co-authored “On the number of partition weights with Kostka multiplicity one” with Z. Gates and R. Vinroot: E lectronic Journal of Combinatorics 19 (2012), no. 4.

● Designed a matrix inversion approximation algorithm 20 times faster and 30% more accurate than commercial software on a special class of matrices.

● Developed a patient clustering procedure estimated to cut mileage driven by in-home caregivers up to 70%.

● Derived a graph-theoretic predictor of social network friendship strength with accuracy competitive to that of popular metrics.

Proficient in Spanish, Literate in French. Citizenship: USA.

Contact this candidate