Name: Mathew Wingerson (he/him)
Location: Denver, CO (relocating to... tbd)
Job: PhD Student in Rehabilitation Science with focus in Data Sciences and Health Analytics
Company: University of Colorado School of Medicine | Colorado Concussion Research Laboratory
Email: Mat.Wingerson@gmail.com | Mathew.Wingerson@CUAnschutz.edu

GitHub (@WingersonMJ) || linkedin || MyBibliography || Posters/Presentations

Link to my full CV: CV Link

I am passionate about identifying patterns in data and using data-driven decisions to effect change in healthcare and health sciences research. I have a strong educational background and significant real-world experience in applied statistics and data sciences, including modern machine learning and deep learning architectures. I enjoy using my knowledge to contribute collaboratively on research that impacts clinical practice and provision of medicine.

My 6 years of experience in clinical research and statistics, and 3+ years of processing and integrating multimodal data sources (text, clinical, electronic, wearable, biologic, imaging), and applying AI and ML techniques (neural networks, clustering, tree-based classifiers, etc.) have prepared me for a position as an AI/ML Health Sciences Researcher.

Expected: 2026
University of Colorado Anschutz School of Medicine, Aurora, CO.

Completed: October 2025.
Conferred: May 2026.
University of Colorado School of Public Health, Aurora, CO.

Conferred: August 2021.
Adams State University, Department of Kinesiology, Alamosa, CO.

Conferred: December 2018.
Simpson College, Department of Psychology, Indianola, IA.

I have a passion for identifying patterns in data and for using those patterns to make predictions and guide clinical management for medically complex injuries! I have a strong foundation in applied mathematics, including modern machine learning and deep learning architectures, and I enjoy using that knowledge to solve real-world problems. I am well-experienced in processing various data types (tabular, image, text), caplitalizing on a variety of feature sources (bench/clinical research, electronic medical records, imaging), and applying a diverse range of AI and ML techniques (Neural Networks, Clustering Algorithms, Tree-Based Classifiers, etc.) to predict health outcomes in children and adults.

I thrive when allowed to leverage conventional statistics, machine learning, and deep learning to make sense out of imperfect data. I am looking for a career that allows me to use these tools to tackle increasingly complex problems in fields where simple solutions fall short!

I am ahead of schedule in my progress toward a dissertation defense in 2026. I am currently searching for opportunities to use my passion and skills in AI and Machine Learning to contribute to a team of researchers looking to make an impact on real-world problems.

Link to my full CV: CV Link

• Git (version control)
• Markdown
• R, R-Studio, and R-Shiny (4+ years)
• Python (3+ years)
• VSCode IDE
• Scikit Learn
• Pytorch and TensorFlow
• Conda/Anaconda (environment and dependency management)
• Docker (just getting started)
• Windows CMD (familir with Bash)
• Snakemake (workflow development)
• Pre-Commit (linting/formatting with Black, Ruff, isort)

Like-Me Clinical Aggregation: A framework for aggregating recovery outcomes among past patients based on similarity matching of clinical characteristics.
Summary: This patients-like-me aggregation approach generates sub-cohorts that better reflect individual clinical presentations and recovery trajectories, offering an interpretable, data-driven complement to clinical management and supporting patient-centered discussions of expected recovery.
Project Repo: https://github.com/wingersonMJ/Like-Me_Clinical_Aggregation

NeurIPS 2025 Submission: Leveraging ordinal embeddings for predicting and generalizing health outcomes: A study of adolescent substance use.
Summary: Deep representation learning for large-scale adolescent health data; introduced ordinal embeddings that outperform and better generalize to hold-out data than one-hot encoded baselines on substance-use risk prediction.
Project Repo: https://github.com/wingersonMJ/2025_NeurIPS_Submission/

Decision Tree: A clinically intuitive approach to predicting persisting symptoms after concussion in adolescents.
Summary: Interpretable decision-tree classifier for clinical use; tuning of hyperparameters, including depth, minimum samples per split, re-weighting for class imbalances, and cost-complexity pruning with selection of alpha value; stratified K-fold validation for model performance; bootstrapping with replacement for confidence intervals; variations of decision-tree visual representations for immediate use in clinical settings.
Project Repo: https://github.com/wingersonMJ/Decision_tree_prediction

PropensityBatchRandomization: A tool for randomizing participant biological samples across plates/batches while numerically evaluating the balance of key participant covariates post-randomization.
Summary: Python package, published on PyPI, open source with tutorial; providing a structured tool for randomizing participant biological samples across plates/batches and numerically evaluating and balancing key participant covariates post-randomization; primary objective to mitigate batch effects encountered in analysis of biologics.
Project Repo: https://github.com/wingersonMJ/PropensityBatchRandomization

Ensemble Learning of Proteomics Prediction Models: This project is the start of model building for a study that uses proteomics to predict curve severity in patients with scoliosis.
Summary: The objective is to take information gained about cellular functions, disease states, and biological pathways obtained from analyses of 7,500 proteins to estimate a patient's Max Cobb angle - the largest point of spinal curvature measured in degrees on an X-ray. I trained multiple neural networks to predict the target variable, then used ensemble learning by training a stacked linear regression on the network outputs. Project Repo: https://github.com/wingersonMJ/Proteomics_Neural_Network_Ensemble

#1. Nirvana	#2. Foo Fighters	#3. Alice in Chains

#4. The White Stripes	#5. Taylor Swift	#6. Jimi Hendrix

#7. Weezer	#8. The Killers	#9. The Offspring