Dmytro Taranovsky

110 Bow St
Lexington, MA 02420 
Email: dmytro@mit.edu

Education

Massachusetts Institute of Technology, Bachelor of Science in Mathematics (Minor in Physics) 2005

Knowledge and Skills

Programming Languages and Frameworks: Python (preferred), NumPy, Bash, Perl, C++, HTML, SQL.
Operating Systems: extensive Linux experience (both command line and Fedora Linux as desktop), knowledge of computer architecture.
Analytical: machine learning, theory of algorithms, computability, complexity.
Mathematics: linear algebra, abstract algebra, statistics, mathematical analysis, topology, mathematical logic, set theory.
Physics: general knowledge, classical mechanics, special and general relativity, statistical physics, quantum mechanics.

Work Experience

eScription (division of Nuance) (August 2006 – June 2011)
Software Engineer
Worked on computer assisted medical transcription (speech recognition and text formatting). Created algorithms and developed software to effectively use very large data amounts, and achieved 15% accuracy improvement. Worked with a large computer farm, participated in research conferences (ICASSP, Nuance Research Conference), and maintained and cleaned up legacy codebases. Used Python, Perl, shell, and C++ on Linux.
• Created a component to efficiently learn and apply millions of text mappings from billions of words of text. Implemented various techniques for best selection and application of the mappings.
• Contributed to various formatting components for finding headings, formatting numerics, numbered list recognition, adding punctuation, applying rule-based text transformations, etc.
• Work involved full software lifecycle: Design, programming, research/experiments, testing (including rollout and production monitoring, and addressing customer support tickets), and software maintenance and bug fixing.

Parametros Trading (December 2012 – January 2015)
Software Engineer
Analyzed performance of our high-frequency stock trading system, measuring patterns in very noisy data and running simulations, and made improvements to the trading algorithm. Analysis of stock executions for a large hedge fund. Engineering work to improve robustness and scalability. Used Python, shell, and C++ on Linux.

Research

2015 – present (and earlier)
Conducted independent research in set theory, mathematical logic, and computer science (see my research statement for additional details):
• Extended the language of set theory and provided semantics and axiomatization for higher order set theory; developed higher order set theory and demonstrated its close connections with large cardinal axioms (old paper and new paper and a further extension).
• Created a strong ordinal notation system (link), and wrote a Python program for ordinal comparison and arithmetic using the system (link).
• Defined arguably finitistic properties of complexity up to the second order arithmetic (link).
• Defined constructive truth for arithmetic and analysis, and developed a corresponding constructive theory (link).
• Proposed a solution to the Generalized Continuum Hypothesis (link1 and link2).
• Defined a finite analogue of infinite models (link).
• Used probability to make conjectures in number theory about representability of numbers as sums of powers and primes (link).
• Proved an equivalence between WKL0 and a determinacy hypothesis (link).
• Paper "Determinacy and Fast-growing Sequences of Turing Degrees" (link).
• Paper "Arithmetic with Limited Exponentiation" (link).
• Paper "Space-Efficient Circuit Evaluation" (link).
• Paper "Elementarily self-embeddable models of ZFC" (link).
• Paper "Asymptotic Optimal Sphericity" (link).
• Proved that sorting is possible with lg(n!) + o(n) comparisons (link).
• Paper "AI Notes" (link) (2023).

Interests and Activities

• Mathematics, especially set theory; computers and computer science, including machine learning and AI; physics; philosophy.
• Reading and studying.
• Contributor to FOM (foundations of mathematics) mailing list, Wikipedia, and StackExchange.
(May 2015 – Sep 2015) Technical reviewer for Python Machine Learning book.

Awards

• Member of the United States Physics Team and participation in the summer training (1999)
• First Place in University Of Maryland High School Mathematics Competition (among high school students in Maryland, 2000)

Languages

English, Russian, Ukrainian