We develop machine learning tools and neurotechnologies to decode brain function, model mental illness, and build precision psychiatry applications. Our research spans graph neural networks, multimodal learning, dimensionality reduction, and large-scale neural data collection and analysis.

• QuantNets
  Generalizing Convolutional Neural Networks (CNNs) to graph data using Learnable Neighborhood Quantization. This approach allows for more accurate and expressive models on graph-structured data. NeurIPS 2024

• EpiCare
  A reinforcement learning benchmark for dynamic treatment regimes in healthcare. It includes simulated episodes of care with realistic reward sparsity, partial observability, and heterogeneous treatment effects. NeurIPS 2024

• [CARVE-AI][coming soon]
  Simple and scalable algorithms for cluster-aware precision medicine. This project focuses on developing methods that enhance precision medicine by incorporating clustering techniques to identify patient subgroups, aiming to improve treatment strategies and healthcare outcomes. AISTATS 2024 Paper

• MultimodalFeatureSync
  A comprehensive pipeline for processing video files to perform speaker diarization and extract a rich set of features encompassing acoustic, linguistic, and facial data.

• CLEAN
  [Coming Soon] Tools for data cleaning and preprocessing pipelines for EEG and clinical data.

• Website: grosenicklab.org
• Bluesky: logangrosenick.bsky.social
• Twitter/X: logangrosenick

Want to contribute? Reach out or open an issue on any project above.