Research interests

I develop probabilistic AI for military and financial applications. My major research foci are

  • Nonparametric and Bayesian structural time series
  • Agent-based modeling
  • Mechanistic models of non-equilibrium systems
  • Probabilistic programming

Application areas in which I’m interested include

  • Logistics and supply chain
  • Cybersecurity
  • Geopolitical risk
  • Enterprise risk management

Description

Software

  • CrimsonSkyline.jl: a universal trace-based probabilistic programming language, hosted in Julia.
  • stsb3 Structural time series, round 3. Restriction and refinement of stsb2 grammar and extension of modeling capabilities, deeply integrated with Pyro and pytorch.
  • stsb2 Structural time series, round 2. Implements a grammar over structural time series models and an associated modeling + inference library.
  • verdantcurve: agent-based market model created in collaboration with Colin Van Oort (MITRE Corporation).
  • discrete-shocklet-transform: qualitative, shape-based, timescale-independent time series similarity search algorithm