Eric Hughes, PhD

Data Science

Specialties

Large-scale Simulations of Physical Systems
Analysis of Geospatial Imagery
Statistics
Numerical Modeling

Education

Ph.D. Atmospheric Science, University of Maryland, College Park
M.S. Physics, UMBC
B.A. Visual Arts, UMBC

Location

Maryland

Recent Projects

Multinational Construction Company
Estimating and Scheduling Mega Construction Projects
Modeled the end-to-end flow of large and complex construction projects. Derived features and models to validate project estimates, leading to tighter estimates and reduced financial contingency requirements.

Civilian Scientific Government Agency
Multivariate Analysis of Satellite Data Biases
Developed a multivariate analysis to discover key atmospheric quantities that produce biases in satellite data. Predicted specific conditions where anomalously errors in satellite data will occur, providing error-correction guidance to satellite data developers. Improved NASA’s ability to monitor volcanic eruptions and understand the effects volcanoes on earth’s climate.

Technical Expertise

  • Spark, SQL, Hadoop, EMR
  • MPI on supercomputer clusters
  • Python data stack: scikit-learn, pandas, numpy, Jupyter, others
  • Scientific analysis, statistical analysis, numerical modeling
  • Machine learning: regression analysis, time-series analysis, Bayesian statistics, data assimilation, inverse problems
  • Deep learning on Tensorflow and pytorch
  • Satellite systems