My research focus is on modeling and solving large-scale optimization problems and on finding and exploiting structure in high dimensional data, with applications in marketing, finance, demographic modeling, and medical informatics.

low rank models

Generalized low rank models (GLRMs) extend the idea behind principal components analysis (PCA) to embed tabular data sets with heterogeneous (numerical, Boolean, categorical, and ordinal) types into a low dimensional space, providing a coherent framework for compressing, denoising, and imputing missing entries.

papers

• long version (the full framework)

• short version (on algorithms)

• short version (on modeling)

• shortest version (on medical applications)

applications

• discovering patient phenotypes using GLRMs (using GLRMs to embed medical records into a low dimensional space)

code

• julia (fully featured, serial or shared memory parallel)

• H2O (fully featured, distributed memory parallel)

• python (prototype, serial)

• spark (prototype, distributed memory parallel)

talks and slides

• full deck

• H2O World video and slides

• H2O world demo (by Anqi Fu)

• INFORMS slides

convex.jl

Convex.jl is a Julia package for Disciplined Convex Programming (DCP). Convex.jl makes it easy to describe optimization problems in a natural, mathematical syntax, and to solve those problems using a variety of different (commercial and open-source) solvers. Convex.jl is one of the top ten packages in the Julia language for technical computing, and is a member of the JuliaOpt organization, which curates high quality optimization software.

paper

• conference paper

code

• source code

• documentation

talks and slides

• JuliaCon 2015 video

• JuliaCon 2014 video and slides