An introduction to Bayesian machine learning through a collection of notebooks
In this notebook, I implemented the Naïve Bayes Classifier and the Bayesian Linear Regression and applyied them to the MNIST and CIFAR10 datasets.
Since a CNN is capable of modeling spatial patterns through the hierarchical analysis of patches of images, it was interesting to apply Naïve Bayes classification to the output of convolutional layer in the LeNet architecture to see the improvement of the accuracy.