Hello, world!
My name is Joe Dehoney
About Me:
I am currently working at Stanford University as a Research Data Analyst. I am also affiliated with SRI International.
Interests:
Machine Learning, Deep Learning, Computer Vision, High Performance Computing, Data Science, Ambient Intelligence, and Medical Imaging Analysis.
Technical Skills:
- Languages: Python, C++, SQL, Bash
- Libraries: Pandas, Numpy, Matplotlib
- AI/ML: PyTorch, Tensorflow, Keras
- Parallelism: CUDA, OpenCL, OpenMP
- Tools: Airflow, AWS, Docker, Git, Jupyter, Kubernetes, Linux, REDCap, MySQL, MongoDB, Slurm, XNAT
Resume/CV:
Click Here to download a copy of my resume
Work Experience:
Research Data Analyst
Computational Neuroscience Laboratory, Department of Psychiatry and Behavioral Sciences, Stanford University (05/2022 - Curr)
Responsibilities:
- Created data submission pipeline (incl. unit testing) to upload 10 years of neuroscience data to NIH repository
- Invented tools for efficiently resolving data quality issues reducing resolution time by 80 percent
- Aggregated data from multi-center studies by leveraging Python to harvest unstructured web content for use by lab researchers
- Unified processing pipeline using containerization to enable deployments across different computing environments
- Collaborate with researchers to curate datasets within REDCap to support longitudinal data analysis, and to facilitate development and evaluation of machine learning models
- Report frequently to government officials and stakeholders about compliance of data collection and distribution
- Administrate linux compute clusters (10 nodes) and storage servers (1 Petabyte)
Product Manager
CoolWine, Santa Rosa CA (07/2019 - 04/2021)
Responsibilities:
- Assisted in design, fabrication, and deployment of industrial beverage coolant systems
- Created the data pipeline and analysis for longitudinal equipment analysis
Student Manager
UC San Diego Recreation, San Diego CA (05/2017 - 06/2019)
Responsibilities:
- Designed and maintained the inventory system for 6 facilities
- Managed a team of 45 undergraduates, directing project creation, delegation, and supervising performance
Team Lead
Permanent Value Inc., San Diego CA (07/2016 - 05/2018)
Responsibilities:
- Managed 60 client investment portfolios totaling over 15 million in assets
- Proposed and implemented investment strategies and financial trades
- Presented recommendations and analysis to the company founder on a weekly basis
Education:
Oregon State University
BS in Computer Science; 01/2022 - 06/2023
GPA: 4.0
Awards: Summa Cum Laude, Dean's List, Engineering Scholarship
UC San Diego
BA in International Business, Minor in History; 09/2015 - 06/2019
GPA: 3.9 (Major), 3.2 (Cumulative)
Awards: Provost Honors
Projects
Deep Reinforcement Learning for Financial Trading
- Created a deep reinforcement learning model based on Deep Q-learning for trading cryptocurrencies.
- Built a comprehensive data pipeline to collect, clean, and preprocess cryptocurrency market data.
- Fine-tuned the model's hyperparameters and architecture through extensive experimentation and analysis.
- Conducted thorough backtesting and performance evaluation to assess the model's effectiveness.
- Integrated the model with cryptocurrency exchanges through API integration for live trading.
- Documented the project's methodology, codebase, and results in a clear and concise manner.
Ecommerce Web Application
- Created with other members of the data science team from the OSU App Development Club an AI system for demand forecasts of products
- Programmed models in PyTorch library.
Publications
Qingyu Zhao, Magdalini Paschali, Joseph Dehoney, et al., Career Path Matters for Identifying High School Risk Factors that Forecast Heavy Drinking in Young Adulthood: Under Review (2023)
Lauren Kupis, Zachary T Goodman, Salome Kornfeld, Stephanie Hoang, Celia Romero, Bryce Dirks, Joseph Dehoney, et al., Brain Dynamics Underlying Cognitive Flexibility Across the Lifespan, Cerebral Cortex, Volume 31, Issue 11, November 2021, Pages 5263–5274