My original Brandeis website:

Blake LeBaron

Short biography for Blake LeBaron

I’m curently a professor at Brandeis University in the International Business School.

My research and teaching are centered on computational modeling both empirical and theoretical.

My primary contributions are in the development of agent-based models for financial markets. These models allow for the interactions of many agents in a market, and concentrate on the interactions of their forecasting models. I have also built agent-based models in other areas as well. A very quick summary is that I’m interested in all areas where outcomes involve the coevolutionary dynamics of populations of predictive models.

I also have been active in time series modeling. In particular dealing with time series forecasting models for financial markets with a special interest in volatility models.

Active research projects:

  • Simple agent-based microstructure models
  • Macroeconomic models with heterogeneous forecasters
  • Long range models for volatility/liquidity dynamics

Recent courses taught:

  • Machine learning
  • Python for business and finance
  • Forecasting
  • Computational risk modeling
  • Advanced asset pricing (Ph.D)

Computer skills/interests

I’m an active programmer, but try not to be super technical. I do try to keep up with things that are happening, and I’m very interested in squeezing the most speed possible from relatively modest hardware. My skills cover:

  • Python: (very active)
    • ML tools for Python (Scikit-learn, Keras/tensorflow, numba)
  • Julia: (I’m a big fan, and trying to shift code to this platform)
  • Netlogo: (I use this for some agent-based models)
  • Matlab: (much of my early work uses this)
  • R: (used for teacing stats/forecasting)
  • Objective-C/C++: (not very often)
  • Mojo: (have started looking into this new language)

In terms of software tools, I’m very interested in talking to anyone in the Julia community. I’m hoping it will start to catch on in computational finance/time series work.