List of topics to cover:

  • Bayesian solution to overfitting
    • Salisman's solution to the Don't Overfit
  • Predictive distributions; "how do I evaluate testing data?"
  • model fitting, BIC + visualization tools
  • Gaussian Processes

Would be nice/cool to cover:

  • classification models (using the books text)
  • Bayesian networks?
In [10]:
from IPython.core.display import HTML
def css_styling():
    styles = open("../styles/custom.css", "r").read()
    return HTML(styles)