The following repository gives an implemntation of a MATLAB tutorial on classifying wafer defects using convolutional neural networks (CNNs): https://www.mathworks.com/help/deeplearning/ug/classify-anomalies-on-wafer-defect-maps-using-deep-learning.html.
In python, the TensorFlow software is used to implement a convolutional network trained on the same training and testing database used for the MATLAB code. The Jupyter notebook shows the results of the CNN model as well as its PR curves and confusion matrices, similar to the MATLAB tutorial above.
Database was created by MIR Labs: http://mirlab.org/dataset/public/.