CHRIS MILES
Data Scientist
Pacifica,CA
Summary
I am a data scientist formally trained in applied mathematics (MS) and physics (PhD) with an expertise in optimization, time-series analysis, and forecasting. I have over 10 years of experience with scientific computing projects primarily in Python and R.
Overview
12
12
years of post-secondary education
7
7
years of professional experience
Work History
Data Scientist
Big Data Federation
Santa Clara , CA
2019.04 - Current
- Forecasted company financial fundamentals such as revenue and earnings per share for publicly traded companies to inform trading strategies using tree-based models and statistical time-series models.
- Developed stock trading strategies by forecasting stock price movements using probabilistic graphical models and genetic algorithms.
- Used random matrix theory methods for detecting signal from noise in the feature selection process and for de-noising techniques.
- Applied quadratic programming optimization methods to portfolio asset allocation.
- Involved with entire machine learning pipeline from data wrangling, data cleaning, modeling, and writing code in production.
- Built R packages for internal tooling which involved implementing automated unit tests and using version control with git.
- Developed visualizations and dashboard using Shiny and ggplot2 libraries to present results to management team.
Data Science Fellow
Metis
San Francisco , CA
2019.01 - 2019.03
- Developed 5 data projects, focused on business-applicable outcomes, as part of a 3-month immersive program covering statistical modeling and machine learning techniques, using Python and SQL.
Graduate Student Researcher
University Of Michigan
Ann Arbor , MI
2012.09 - 2018.05
- Discovered how to efficiently mix a dye within a fluid by choosing optimal fluid stirring strategies through optimal control theory approaches which contributed to the broader understanding of fundamental fluid processes.
- Built fluid simulations from scratch in numpy which involved implementing finite difference and spectral methods for solving fluid equations.
- C. J. Miles, C. R. Doering, Diffusion-limited mixing by incompressible flows, Vol 31, Issue 5, Nonlinearity, 2018.
Education
Ph.D. - Physics
University Of Michigan
Ann Arbor 2012.09 - 2018.05
Master of Science - Applied Mathematics
University Of Michigan
Ann Arbor 2012.09 - 2015.01
Bachelor of Science - Physics
Massachusetts Institute Of Technology
Cambridge, MA 2006.09 - 2010.05
Skills
Python
undefinedPersonal Projects
Kaggle Competition: Conway's Reverse Game of Life (placed top 4%)
- Goal is to generate a starting state that when evolved forward in time, according the rules of John Conway's Game of Life, closely matches the given final state — thus reversing the game of life in a sense.
- Created a model with an average cell accuracy of 96.5% of the evolved final state from proposed starting states across 10,000 games. Model placed in top 4% on public leaderboard with over 180 competitors.
- Used simulated annealing optimization with the help of GPU computation in pytorch to parallelize the evolution of a thousand games simultaneously at each iteration of algorithm.
Timeline
Data Scientist
Big Data Federation
2019.04 - Current
Data Science Fellow
Metis
2019.01 - 2019.03
Graduate Student Researcher
University Of Michigan
2012.09 - 2018.05
Ph.D. - Physics
University Of Michigan
2012.09 - 2018.05
Master of Science - Applied Mathematics
University Of Michigan
2012.09 - 2015.01
Bachelor of Science - Physics
Massachusetts Institute Of Technology
2006.09 - 2010.05