What we will do?

Discuss the role of Bayes theorem for hyper-parameter optimization of deep-neural networks. Bayes' theorem was created by: Thomas Bayes, a mathematician & a Presbyterian minister. The theorem was popularized posthumously by his friend Richard Price.

The theorem, alternately referred to as Bayes-Price theorem, to reflect the significant contributions Price made to the original work of Bayes, is a mathematical formulation to update beliefs.

The original work of Bayes & Price was independently extended by French mathematician Pierre-Simon Laplace, who published the modern formulation of the theorem.

Due to the usefulness of Bayes' theorem on how to update beliefs, it is a powerful tool for hyper-parameter tuning in deep-neural networks.

This workshop will go through the basics of Bayes' theorem & mechanism of optimizing a deep-neural network using Bayesian strategies. We will be using:
1) Tensorflow & Keras
2) SciKit, SciKitOpt, GPy & GPyOpt

Minimum requirements:

Understanding of Python 3 & basics of probability theory, is recommended, not necessary.

Important to know:

This is a ticketed event. Tickets are available here: https://www.moad.computer/store/p36/Healthcare_Analytics.html
Please bring a valid ID for the front desk.
This is a bring your won device (BYOD) event. Please bring a laptop with Google Chrome (>=71) installed to effectively follow along the content.