3.1.2.9. Analyzing machine learning model testing results

This page briefly summarizes the contents of the Jupyter notebook at the file path <root>/examples/modelling/cbed/distortion/estimation/notebooks/analyzing_ml_model_testing_results.ipynb, where <root> is the root of the emicroml repository.

In this notebook, we analyze the output that results from performing the “actions” described in the following pages:

1. Generating simulated CBED patterns of a sample of MoS2 on amorphous C

2. Generating machine learning datasets for the machine learning model test set #1

3. Combining machine learning datasets for the machine learning model test set #1

4. Running the machine learning model test set #1

5. Running the RGM test set #1

In short, in this notebook we analyze the performance results of the “first” set of machine learning (ML) model tests for the ML task of estimating distortion in convergent beam electron diffraction (CBED). These performance results are benchmarked against those obtained by the radial gradient maximization (RGM) approach to estimating distortion.

In order to execute the cells in this notebook, a set of Python libraries need to be installed in the Python environment within which the cells of the notebook are to be executed. See this page for instructions on how to do so.

It is recommended that you consult the documentation of the emicroml library as you explore the notebook. Moreover, users should execute the cells in the order that they appear, i.e. from top to bottom, as some cells reference variables that are set in other cells above them.