ML/AI
Projects

Production-grade machine learning systems bridging computational physics and artificial intelligence. Interactive demos, open-source code, and deployable models.

Featured ML/AI Projects

Each project includes source code, interactive demos, and documentation.

ML/AI Competencies

Deep Learning

  • Physics-Informed Neural Networks (PINNs)
  • Residual Networks with skip connections
  • Multi-task learning architectures
  • Feature attention mechanisms
  • Focal loss for class imbalance

Frameworks & Tools

  • PyTorch (custom losses, training loops)
  • TensorFlow & Keras
  • Streamlit (interactive dashboards)
  • Plotly & Matplotlib (visualization)
  • NumPy, SciPy, Pandas

Scientific Computing

  • Plane Wave Expansion (PWE) method
  • Eigenvalue solvers (Jacobi, QR)
  • Latin Hypercube Sampling (LHS)
  • Berry phase computation
  • Dispersion relation analysis

Engineering Practices

  • Modular architecture with type hints
  • Comprehensive test suites (pytest)
  • CI/CD pipelines (GitHub Actions)
  • Early stopping & LR scheduling
  • Production-ready deployment
← All Projects Publications → GitHub Profile