A colorful illustration of a data flow or network diagram with nodes and connecting lines, using shades of orange, blue, purple, and yellow.

Accelerate your research.
Advance human health.

MATCH with AIM-AHEAD experts and AI tools

Two speech bubbles, one blue with sparkles and squiggly lines, and one orange with a large question mark.
Click for MATCH AI Chatbot

Expand your team’s AI proficiency

Illustration of a video player with a cartoon-style gold medal or award on a stand, and some handwritten-style black lines on a teal background.
Enroll in the AIM-AHEAD and MATCH AI Hands-on Workshop Series on AIM-AHEAD Connect
Access our workshop materials on our GitHub
Logo featuring a star with a compass and a globe, with the words 'AIM AHEAD'.

Part of the AIM-AHEAD Data and Infrastructure Capacity Building (DICB) Program

MATCH aims to ignite & accelerate AI and machine learning projects in biomedicine for AIM-AHEAD awardees and their teams

Workshops

Upcoming

  • This tutorial will introduce the core concepts of Continual Learning, focusing on how models can learn incrementally without forgetting previous knowledge. Participants will engage in hands-on exercises to implement key continual learning techniques using PyTorch. The session will also cover practical applications in AI and healthcare, demonstrating the importance of continual learning in real-world scenarios.

Jayanta Dey, Ph.D.
Post-doctoral research fellow @ UTSA MATRIX AI Consortium

Workshop materials & information

  • Increasingly, generative AI models have shown a lot of promise in a variety of biomedical applications, such as Alpha-Fold for generating sequences of proteins and Evo for generating DNA sequences. This workshop gives an overview of current state-of-the-art methods for biomedical applications and how users can get started using code.

Sambit Panda, Ph.D.
AI Research Scientist @ UTSA MATRIX AI Consortium

Christian Cruz
AI Engineer @ UTSA MATRIX AI Consortium

Workshop materials & information

  • This module introduces beginners and early-stage AI practitioners to predictive modeling through hands-on coding exercises. Using a real-world acute myeloid leukemia dataset of 100 patients, learners build models to predict patient responses to treatment. The module begins with core concepts in model development and standard machine learning pipelines, then covers linear and logistic regression, decision trees, and random forests. The interpretability of models is assessed using Shapley Additive Explanations (SHAP) for logistic regression and feature importances for random forests. The module concludes with optional self-guided projects that build on and improve the techniques introduced.

Mignon Frances Dumanjog, M.S.
Research Software Developer @ UTSA MATRIX AI Consortium

Computer vision for cell confocal microscopy

David Hernandez-Guzman
PhD Candidate @ UTSA Biomedical Engineering

Mignon Frances Dumanjog, M.S.
Research Software Developer @ UTSA MATRIX AI Consortium

  • In this module of the MATCH Workshop Series, students will learn how to prepare and clean claims data, define spatial relationships, and compute measures of spatial dependence. The module emphasizes both theoretical understanding and practical implementation using R and RStudio, guiding learners through preprocessing workflows, spatial weights construction, and interpretation of clustering patterns in healthcare outcomes. Hands-on exercises provide experience applying global and local Moran’s I to real-world datasets, while discussions highlight common pitfalls, such as the Modifiable Areal Unit Problem (MAUP) and data quality issues, ensuring critical evaluation of results.

    By the end of this module, students will be able to:

    1. Define geospatial analysis and explain the role of spatial autocorrelation in healthcare and public health research

    2. Differentiate between global and local measures of spatial autocorrelation, and interpret the outcomes of Moran’s I

    3. Execute a complete Moran’s I workflow by preparing data, defining spatial neighbors, calculating the statistic, and interpreting results

    4. Preprocess healthcare claims data, construct spatial weights matrices, and implement Moran’s I analysis using appropriate libraries using R & RStudio.

Paula Andrea Jaimes-Buitron
Ph.D. Candidate @ UTSA Biomedical Engineering

Workshop materials & information

Past workshops

Core areas of expertise

  • Illuminated speech bubble sign

    Fine-tuning LLMs

  • Line graph showing multiple colorful trend lines with upward movement on a dark background, related to data analytics or stock performance.

    Forecasting / prediction models

  • A digital data wall with blue lights and lines representing data streams or server racks.

    Federated learning

  • Abstract digital network with interconnected glowing blue lines and nodes forming a helix shape.

    Continual learning

  • Close-up of colorful LEGO bricks arranged in a pattern.

    Computer vision

  • Colorful glow-in-the-dark jellybeans against a black background.

    Molecular & omics data analysis

  • Close-up of a spreadsheet or data sheet with rows of numerical data and decimal numbers.

    Model benchmarking and scoring

  • Nighttime aerial view of a city with illuminated roads and highways.

    Geospatial population-level models