DIAMOND
Notebooks
EN

ML NOTEBOOKS

A notebook whose output cell compares predictions with actual values, a decision tree whose path leads to a prediction, and an autoencoder

Machine learning is part of DIAMOND’s expertise for the PEPR DIADEM community. This section gathers the resources that put it into practice on materials data: educational notebooks you can run and adapt to your own datasets, and a pipeline developed within the project and applied to experimental characterization.

They serve two different purposes. The training notebooks are there to be followed, modified and transposed: they walk through a complete workflow, from checking the quality of a dataset to building predictive models and exploring more advanced approaches. The Raman denoising pipeline is a production tool, published with its article and its data, that you can reuse on your own spectra.

Available resources

  • ML training notebooks

    A step-by-step path applying ML methods to materials data, from data quality to Bayesian optimisation

  • Raman denoising pipeline

    Noise2Noise denoising pipeline for high-throughput Raman spectroscopy

And more to be added!