Michael D.

Experience Summary

My diverse skills and experience in mathematics, statistics, and computer programming should be advantageous in this circumstance. I have worked with mathematical software and statistical models since 1989 and love doing collaborative statistical and mathematical research to solve challenging problems. I hold a PhD in Mathematics and a Master’s Degree in Statistics. I recently worked as an independent machine learning consultant. In this capacity, I dealt routinely with HMM (Hidden Markov) models, neural networks, and LSA and LDA analysis of text data. This work consisted of programming in C++, Perl and Python and also required mastery of mathematical theory such as forecasting, Markov chains, stochastic processes, Bayesian statistics, and operators on Hilbert space. One of the tasks I accomplished was to improve and then implement in C++ an algorithm which used a projected subgradient method to simultaneously learn several linear classifiers (perceptrons) using as few features as possible (“transfer learning”). I am also involved in developing software to find posterior distributions (of forecasted quantities) using Gibbs sampler (MCMC) methods. Most of this recent work is being implemented in a Linux environment (using Apache Subversion/SVN as the version control system). Recently, at Premier Research Group (CRO), I developed software in C++ to implement statistical methods for adaptive designs in clinical trials in both the frequentist and Bayesian frameworks. (I have a total of about 15 years of experience programming in C++.) I also managed and mentored other statisticians and programmers in the Biostatistics department and supervised output quality control. This involved writing and testing various SAS programs. (I have about 20 years of experience using SAS to compute inference from statistical models.) Finally, I created client presentations concerning various adaptive trial designs and regularly wrote/revised protocols in response to FDA special protocol assessments. I have the equivalent of about 10 years of experience using R and S-plus to perform calculations with data, especially those involving linear algebra. I have also often had occasion to utilize the graphics capabilities of S-plus for many projects. As a consultant at Yale (School of Nursing), I did extensive data manipulation with Perl, performed analysis of various outcome data, and wrote computer programs to do power and sample size calculations in C++ and Java. Much of the work at Yale involved analyzing longitudinal data using mixed regression models that took into account autocorrelation over time (proc MIXED in SAS). At Tufts, I collaborated with Vanderbilt University researchers to investigate linkage analysis of various regions of the genome related to autism (statistical genetics). I used Perl to organize a database linking genomic and nutritional information. I also maintained databases in a UNIX environment, and conducted system administration on Sun Work-stations. I am extremely excited about the possibility of tutoring you. Thank you very much for your consideration.

Teaching Style

Laid back.

My Hobbies

Chess. Guitar.

Other Comments

Proficient in Perl and C++.

All Subjects I Tutor

• Algebra
• Algebra I
• Algebra II
• Algebra III
• C++

• Calculus I
• Calculus II
• College Algebra
• Computer Programming
• Geometry

• Logic
• Math
• Microsoft Excel
• Perl
• Pre-Algebra

• Pre-Calculus
• Python
• Statistics
• Statistics SPSS
• Trigonometry