Bill Chen

A new chapter!

I am Bill Chen, graduated from the University of Kentucky focusing in bioinformatics PhD and Statistics MA, passionate about Big Data, Machine Learning and AI research, with strong interpersonal skills, adept at working in teams and successfully delivering projects. I was awarded Microsoft Azure Research Funding in 2016 and a Grow with Google Challenge Scholarship in 2018. I am a SAS certified advanced programmer and NVidia certified Deep Learning and Cuda instructor.

Experience

Data Scientist / Machine Learning Engineer

Verb Surgical

  • Worked with a multi-disciplinary team to develop surgical analytics software for a digital surgery platform.
  • Improved model performance of more than 5x as measured by accuracy and recall by integrating a video frame data-filtering pipeline and a two-output transfer learning model with CNN and LSTM.
  • Archived a real-time prediction by integrating the signal process methods.
  • Leveraged knowledge in data science, machine learning, statistics, and model scalability.
  • Technologies: Python, R, Computer Vision, PyTorch, Unit-test, CNN, LSTM, VAE, Docker, etc..




    2019 - Present

    University Ambassadors / Deep Learning Institute (DLI) Certified Instructor

    NVidia Deep Learning Institute

  • Delivered training to developers, data scientists, and research on how to use artificial intelligence and GPU-accelerated computing to solve real-work problems across a wide range of domain.
  • Organized and taught workshops on CUDA programming, Fundamentals of Computer Vision (CV), Fundamentals of Deep Learning for Multiple Data Types (MDT), and Fundamentals of Deep Learning for Natural Language Processing (NLP).
  • Provided helps to attendees on training, optimizing and deploying neural networks using the latest tools, frameworks, and techniques for deep learning.
  • Involved in organizing GPU Technology Conference DLI events.



  • 2018 - Present

    Research Collaborator

    Dept. of Statistics, UK

  • Build a High-performance Cluster (HPC) simulation pipeline for the Mix-Gamma Model with R.
  • Simulated data from different gamma distributions.
  • Implemented unit-test, libraries, and workflow for experiments
  • Technologies: SAS, R, R Package Writing, HPC management, Bash, etc..



    2017 - 2019

    Graduate Research Assistant

    University of Kentucky

  • Worked on the construction of Protein NMR Reference Correction and Protein NMR Deuteration Level Detection frameworks.
  • Published Protein Nuclear Magnetic Resonance (NMR) Reference Correction (paper), BaMORC: Bayesian Model Optimized Reference Correction Method for Assigned and Unassigned Protein NMR Spectra (Package) and BMRBr (Package).
  • Built a statistical base model for an estimate of reference correcting values for protein.
  • Implemented a Bayesian probabilistic framework to improve the model performance
  • Surpassed the state-of-the-art performance as measured by reference error below +/- 0.22 ppm at 90% confidence interval. (State of the art is around 1ppm.)
  • Used Python, R, Multi-processing Programming, Statistical Learning, Bayesian, etc.
  • Technologies: R, R Package Creation, Python, Shiny, Docker, etc..



    2013 - 2019

    Profilio

    BaMORC: NMR Reference Correction

    CRAN link · GitHub link · Documentation link.

    BaMORCThe BaMORC package is designed to facilitate protein NMR research with an easy-to-use tool that detects and corrects 13C referencing errors before and after the protein resonance assignment step.






    BMRBr: BMRB Files Downloader

    CRAN link · GitHub link · Documentation link.

    BMRBr is a package that facilites R users to analyze data from BMRB data repo by simplifing the download procedure. Currently, the only way to download individual BMRB NMR-star file is to download manually or using shell code, this package frees R users by allowing users to enter only ID and store location.






    Publication

    Education

    M.A. (Cert) in Statistics

    University of Kentucky

    GPA: 3.57

    Courses include: Statistical Analysis, Design and Analysis of Experiments, Computational Inference, Theory of Probability, Intro to Statistical Methods, Regression & Correlation, Statistical Inference, Clinical Trial, Survival and Life Testing, Linear Model & Experimental Design, Longitudinal Data Analysis, Analysis of Categorical Data.

    2016 - 2019

    Ph. D. in Bioinformatic/Biochemistry

    University of Kentucky

    Dissertation Topic: Automatic 13C Chemical Shift Reference Correction of Protein NMR Spectral Data Using Data Mining and Bayesian Statistical Modeling

    GPA: 3.57

    Courses include: Structural Biology, Biochemistry, Cellullar Biology, Structure & Function of Proteins/Enzymes, etc..

    2013 - 2019

    Computer Science Training

    University of Kentucky

    Courses include: Machine Learning, Computer Vision, Advanced Data Science, Interactive Machine Learning, Numerical Analysis, Calculus and Linear Algebra.



    2016 - 2019

    Certification

    Skills

    Tools



    Programming Languages:

    Python · R · SAS · Shiny · C++ · JavaScript · HTML + CSS · Bash




    Programming Projects

    Interests

    I have undertaken considerable volunteer work, but the most rewarding act of service has been the sponsoring of two children from Romania and Armenia for the past six years. I communicate with these students frequently, encouraging them, and making them feel that there’s someone in their corner.

    I am an active member of ACM, AAAI and SIAM. I was the Vice-President and Treasurer of my department student organization. My roles in this included recruiting members, organizing social events and promoting a healthy and energetic learning environment for ever member.

    I am also a runner and violinist. (I don't play violin while I am running!)

    Awards