Michael Naaman’s areas of expertise include econometrics, statistics, mathematics, industrial organization, machine learning, labor economics, and big data analysis. He has provided econometric analysis supporting complex litigation in major antitrust cases involving alleged price fixing and wage suppression. In a recently published paper, Michael developed “almost sure hypothesis testing.” This new method resolves a paradox in the statistics field by ensuring the probability of a Type I error decreases as sample size grows. He has also developed novel optimization techniques for discontinuous functions, which will be presented at the 2017 World Congress of Global Optimization, hosted by the Texas A&M Energy Institute. Michael is currently working on a paper describing a new statistical distribution, called the volcano distribution, to model stock market returns. In previous work, he developed econometric models for forecasting company sales, and machine-learning models while competing in the Netflix competition. He has also modeled dynamic labor supply decisions over the life cycle. Michael is experienced using a variety of statistical software packages, including EViews, Gauss, MATLAB, R, SAS, Stata, and SPSS.
December 17, 2014 - Dr. Naaman and Dr. Flamm demonstrate that so-called sub-regressions can be statistically unreliable when based on too small subsamples in antitrust class certification.