Skip to Content
Data-driven Material Design

Data-driven Discovery of Optimized Multifunctional Material Systems

Menu
  • About
  • Research
  • People
  • Data and tools
  • Accomplishments
  • News
  • Events
  • Home

Data and tools

Some data and tools available for researchers include:

  • Unsupervised heuristic-based batched dynamic algorithm for pixel-wise image sampling

  • EBSD data generated using DREAM.3D

  • PyTorch library for auto-differentiable EBSD analysis

  • Custom SPPARKS code used to run “candidate grain” simulations of abnormal grain growth

  • Code for actually running “candidate grain” simulations

  • Process for deploying high-throughput experiments with MLflow and Docke

  • Code for running computer vision and graph neural network experiments for predicting AGG in these simulations

Collaborate

We encourage researchers affiliated with the center to submit additional data and tools to Michael Bockstaller.

Data-driven Material Design
5000 Forbes Avenue
Pittsburgh, PA
2020 Carnegie Mellon University / Legal 2025 Carnegie Mellon University / Legal