Bio
I am a data scientist. I used to do web development (fullstack) and system administration for a living.
I have a Bachelor of Science degree in Computer Science and a Master degree in Applied Statistic.
I have over 12 years of professional working experiences as a programmer and over 5 years and counting as a statistician.
My resume.
Currently doing data science and being happy.
The theme is Theme Stack designed by Jimmy.
I hope you have an awesome day.
Education
- California State University, Long Beach Long Beach, CA - Master of Science in Applied Statistics (Biostatistic focus) GPA: 3.727 Cum Laude
- University of California, Riverside Riverside, CA - Bachelor of Science in Computer Science
Thesis
- Tree-based Ensemble Classification Algorithm for High Dimensional Data - CSULB (Master of Science in Applied Statistics)
Publications
- Nicholas Keisuke Brownell, Evan Shannon, Quoc Doan, Gregg C. Fonarow, and Boback Ziaeian (2023 Mar, 81). Healthcare Utilization for US Veterans with Heart Failure during the Covid19 Pandemic (Vol. 31 No. 8). J Am Coll Cardiol. 2023 Mar, 81 (8 Supplement) 536. (https://www.jacc.org/doi/full/10.1016/S0735-1097%2823%2900980-4)
- Pending authors . Pending Title - My Health Veteran Online Utilization and Comorbidities 2023
Employment History
See my resume.
Portfolio
U.S. Department of Veterans Affairs (GLA) (2020 - 2023)
Identifying Racial Healthcare Disparities in Diabetes and Heart Disease
- Analyzed racial and ethnic disparities in healthcare outcomes among minority and women veterans, focusing on diabetes and cardiovascular disease.
- Monitored key health indicators, including blood sugar levels, blood pressure, and medication usage, to assess treatment effectiveness across different racial groups.
- Conducted data stratification by race and ethnicity, addressing challenges such as small sample sizes and multiracial classifications.
- Performed data exploration, cleaning, and covariate creation to ensure robust statistical analysis and modeling.
- Developed data visualizations and dashboards using SQL and R to provide insights into healthcare inequalities and inform clinical decision-making.
- Supported predictive analytics and clinical informatics efforts aimed at improving healthcare equity and optimizing treatment strategies.
- Collaborated with clinical teams and researchers to translate findings into actionable recommendations for policy and program development.
Extracting Homelessness Duration Using NLP from Unstructured Text
- Developed Natural Language Processing (NLP) solutions to extract homelessness and housing duration from unstructured field reports of homeless veterans.
- Transformed qualitative text data into structured numerical formats to enable further research and predictive analytics.
- Applied text mining and entity recognition techniques to identify and standardize temporal information related to homelessness duration.
- Conducted data preprocessing and cleaning to improve text extraction accuracy and model performance.
- Created data models and visualizations to support analysis and inform policy decisions related to veteran homelessness.
- Ensured data integrity and consistency, contributing to clinical informatics efforts aimed at optimizing housing and support services for veterans.
- Supported interdisciplinary collaboration with research teams, helping to enhance the understanding of homelessness trends and improve program effectiveness.
Evaluating Healthcare Effectiveness for Women Veterans (Triage Program – Survey Data)
- Conducted exploratory data analysis (EDA) to assess the effectiveness of a new triage-based healthcare program for women veterans across three healthcare sites.
- Identified and addressed outliers in survey data, ensuring data integrity and study focus by appropriately handling cases such as transgender participants.
- Led the statistical analysis for the project, including data cleaning, visualization, modeling, and summary statistics, to quantify program effectiveness.
- Developed data-driven insights to support clinical decision-making and optimize healthcare delivery for women veterans.
- Created visual reports and dashboards to track program outcomes and enhance the understanding of patient enrollment trends.
- Collaborated with domain experts and clinical teams to translate complex statistical findings into actionable recommendations for improving nursing methodologies and patient care.
- Addressed challenges posed by pandemic-related disruptions, ensuring continuous evaluation and refinement of the triage system.
Women Veterans Whole Health Dashboard
- Developed interactive dashboards in Power BI to track key metrics and outcomes for a pilot program focused on women veterans’ whole health, including stress management, meditation, yoga, and weight loss.
- Conducted data cleaning, transformation, and integration to ensure accuracy and consistency in program analytics.
- Designed data visualizations and reporting solutions to support clinical decision-making and program evaluation.
- Developed performance metrics to assess program effectiveness and optimize resource allocation.
- Acted as a liaison with third-party provider Optum, managing data ingestion and transformation processes through monthly meetings.
- Applied advanced analytics and clinical informatics expertise to support data-driven improvements in veteran health outcomes.
- Ensured seamless integration of structured and unstructured data to provide actionable insights for program development and refinement.
Impact of the Pandemic on Cardiovascular Disease Medication Uptake
- Developed a cohort study based on defined inclusion and exclusion criteria, covering a three-year study period (pre-pandemic, transition, and pandemic periods).
- Extracted and cleaned structured and unstructured data from electronic health records (EHRs) and databases related to cardiovascular medications.
- Conducted exploratory data analysis (EDA) and developed data visualizations to support insights for research publication.
- Analyzed both outpatient and inpatient clinical data to assess medication uptake trends.
- Applied advanced analytics and data modeling to evaluate the impact of the pandemic on medication adherence and patient outcomes.
Cohort Creation for Dye Usage in Body Scans and Potential Cancer Risks
- Developed a longitudinal cohort study analyzing 20 years of veteran healthcare data to investigate potential cancer risks associated with contrast dye usage in body scans.
- Utilized the OMOP Common Data Model for the systematic analysis of observational databases, ensuring interoperability and consistency across datasets.
- Tracked patient data over two decades using inpatient and outpatient records, applying rigorous inclusion and exclusion criteria.
- Integrated and analyzed multiple datasets, including medical procedures, visit types, demographics, pharmaceutical drug data (mail-order, inpatient, outpatient, in-network, and self-reported), and mortality records.
- Conducted data extraction, cleaning, and transformation to support advanced clinical data modeling and predictive analytics.
Mental Health and Wellness Among Veterans Using My HealtheVet
- Conducted clinical data analysis to evaluate the relationship between mental health conditions and substance use disorders among veterans using the My HealtheVet platform.
- Utilized CART (Classification and Regression Tree) analysis to identify correlations (not causation) between mental health status and substance use patterns.
- Created visualizations of CART tree diagrams to illustrate correlations between different covariates, providing clear insights into key factors affecting veteran health.
- Applied predictive analytics and data modeling to assess trends and inform strategies for improving veteran engagement and health outcomes.
- Used findings to support the optimization of My HealtheVet, enhancing its ability to facilitate mental health monitoring, outreach, and engagement.
- Provided data-driven insights to improve clinical decision support models and drive enhancements to veteran healthcare accessibility and wellness tracking.
Fumigate www.fumigatedb.com
Version 0.1.0 Summer 2018 - 2021
I created this website as an authoritative website for perfume. Technology stack is Phoenix web framework. Bootstrap for responsive design and SEO friendly. PostgreSQL for data storage. Server is on Digital Ocean VPS. Ubuntu 18.04 is the OS and was hardened by me. Data is web scraped using scrapy web scrape framework.
Predictive Model for Detecting Toxicity in Drugs
FDA - Food & Drug Administration - National Center Toxicological Research 2017
Implemented non-parametric Bayesian hierarchical model using Chinese restaurant process in R to predict drug induced liver injuries (DILI). This model was going to be use to predict new and old drugs on the market to see if there is a possibility of toxicity base on other known drugs with similar chemical profile. Nonparametric Bayesian is used is used so that the model can extend to multiple data sources, beyond DILI, to make prediction.
Le Snob www.lesnob.com
2014
Fashion website that sell high end luxury fashion for Le Snob brand. Web designer designed the website. I was responsible for translating it into code (PHP, CSS, HTML, Javascript) and for deployment.
Bluepromocode www.dealspotr.com
2013
Redesigned Bluepromocode website using Scala, Ember.js, and Bootstrap 3. Website is renamed to Dealspotr.
Stylespotter www.stylespotter.com
2013
Helped create spiders to web crawl fashion websites and then upload the scraped data onto MongoDB. Scrapy was the webscraping framework used.
PaeDae API www.github.com/PaeDae/paedae-unity-wrapper
2013
PaeDae’s API Unity Wrapper enabling Unity3D applications to integrate with PaeDae web services and serve advertisements.
Baublebox www.baublebox.com
2013
Reviewed PHP’s code, most of the code was on FuelPHP MVC framework. Gave suggestions on refactoring existing code and how to use MVC framework.
UCR Extension Memorandum of Understanding mou.extension.ucr.edu
2012
The whole website code is on the Zend MVC framework using PHP language. There is an internal web tool to control the layout and links.
UCR Extension Program Fee www.iep.ucr.edu/dates_and_fees.php
2011
This is a tool to give potential international students an idea how much studying aboard with UCR cost. I coded this using Zend MVC framework, jQuery, and SQL. Some of the design changed especially the header but my overall code is still intact. The SQL table and schemas were created and design by me.