Daniel Gurevich Email & Phone Number

Los Angeles, Ca, Us

Princeton, New Jersey, United States

MAE 433 (Automatic Control Systems): Instructed and graded a class of 82 students in labs & office hours.

Princeton, New Jersey, United States

I have worked on two distinct projects during my time at Princeton:Sparse physics-informed discovery of physical relations - Formulated a general physics-informed sparse regression framework for automatically identifying partial differential equations satisfied by experimental or simulated data representing either continuous media or systems of discrete particles. Designed multi-level formal languages representing different families of physics models and employed tools from programming language theory to implement software for automated synthesis and symbolic manipulation of data-derived models.Learning on the fly - Identified a provably optimal online control strategy for a general system with unknown linear dynamics exhibiting noise and showed that this strategy achieves a mean performance only slightly worse than the optimal strategy having advance knowledge of the dynamics.

Princeton, New Jersey, United States

Performed research combining data science and natural language processing to enable AI valuation of personal injury cases in the United States.Oversaw rollout of a commercial website enabling affordable access to justice for individuals pursuing PI lawsuits.

Moline, Illinois, United States

Created survival analysis and visualization tool for service parts forecasting using Python lifelines library. Automated non-parametric and parametric fitting of part reliability functions to estimate risk of rare failure events and create recommendations on parts shipments.Benchmarked the state of parts forecasting across the enterprise and formulated strategy for improving parts forecasting infrastructure.

Atlanta, Ga

(with Prof. Flavio Fenton)Developed, implemented, and validated a fast and robust image processing algorithm for tracking centers of activity in optical and electrical recordings of cardiac arrhythmias with up to 100+ million data points and signal-to-noise ratios close to one.Analyzed multi-electrode clinical data, developed classification for mechanisms of fibrillation and compared ex vivo experimental and numerical simulations. Presented results at four international conferences, first author on two peer-reviewed papers.

(with Prof. Grigoriev)Investigated reconstruction of partial differential equation models using sparse regression. Developed highly tunable algorithm achieving 10,000x improvement in accuracy under simulated experimental conditions in comparison to previously published methods. Used numerical analysis to derive theoretical scaling laws and optimize parameters.Presented results at two international conferences, authored two peer-reviewed papers.

Instructed a class of 51 students on Simulation Analysis & Design during and outside of office hours on course material.Graded weekly homework and examinations.

(with Prof. Santanu Dey)Formulated mixed-integer linear programming and conic quadratic models for training of small feedforward neural networks, compared with stochastic gradient descent and Adam.

Anchored the column What’s the Question? in the bimonthly national chess magazine Chess Life Kids, which has over 10,000 readers.

Doctor Of Philosophy - Phd, Computational And Applied Mathematics

Bachelor'S Degree, Triple Major In Industrial & Systems Engineering, Physics, And Mathematics, 4.00 Gpa

Activities and Societies: Alpha Pi Mu Industrial Engineering Honor Society; President of Georgia Tech Chess Club; Co-founder of Student.