SmartTensors

Unsupervised Machine Learning,Supervised Machine Learning,Physics-Informed Machine Learning,Science-Informed Machine Learning,Matrix Factorization,Tensor Factorization,Tensor Networks

SmartTensors-logo
lanl-logo
RD100Awards

SmartTensors is one of the 2021 finalists for R&D100 awards

SmartTensors

SmartTensors is a general high-performance Unsupervised, Supervised and Physics-Informed Machine Learning and Artificial Intelligence (ML/AI).

SmartTensors includes a series of alternative ML/AI methods / algorithms (NMFk, NTFk, NTTk, SVR, etc.) coupled with constraints (sparsity, nonnegativity, physics, etc.).

SmartTensors is developed in Julia.

nmfk-logo
Nonnegative Matrix Factorization

NMFk is a novel unsupervised ML method based on Matrix Decomposition.

ntfk-logo
Nonnegative Tensor Factorization

NTFk is a novel unsupervised ML method based on Tensor Decomposition.

Applications

Research

Resources

SmartTenosrs is open source and available on GitHub

SmartTensors resources include:

  • Codes
  • Scripts
  • Unit tests
  • Test problems
  • Examples
  • Real-world applications and projects (e.g., GeoThermalCloud, ML4Geo, etc.)
  • Jupyter and Pluto notebooks
  • Documentation
  • Videos
  • Tutorials

For more information: info@smarttensors.com tensors@lanl.gov

Velimir V Vesselinov (monty): LANL GitLab GitHub
SmartTensors: Web LANL GitHub Julia
MADS: LANL GitLab GitHub C Julia Python
WELLS: LANL GitLab C Gitlab Julia GitHub
ChroTran: LANL GitHub Gitlab