Thrust 1: Machine Learning (ML) Workflows for Moire Materials

Objective

Discover new correlated and topological moiré systems using advanced machine learning (ML) workflows.

Key Components

• Data Preparation:
  • Aggregating datasets from simulations and experiments.
  • Feature engineering to capture relevant physical properties.
• ML Models:
  • Neural networks for property prediction.
  • Clustering algorithms to identify novel correlated systems.
• Evaluation:
  • Metrics for assessing model accuracy and relevance to physical phenomena.

Tools and Resources

• Python libraries: TensorFlow, PyTorch, scikit-learn.
• Workflow automation with Snakemake.

Example Workflow

1. Data ingestion from high-throughput simulations.
2. Model training on physical property datasets.
3. Visualization of topological phases.

For detailed instructions, visit the HeteroFAM ML GitHub Wiki.

Return to Manual Home