ML Introduction Part 1/2: Theory

Before you dive into some of the hot ML algorithms, it’s important for you to know the basics of whats going underneath the hood — ML is really just fancy optimization. We will start with a refresher on linear regression and represent regression as an “optimization” problem. Next, we’ll show how machine learning is a fancy term for model fitting. A model is a function with parameters and variables. “Fitting” is the process of picking the right parameter values. I’ll show you how sklearn can be used to fit a model. Next, I’ll introduce Neural Networks and how we can represent linear regression as a NN.

  • tags: optimization, math, theory, machine learning, linear regression, neural network, parameters, model fitting, sklearn

Sunday Oct 18th, 1:00am to 2:15am (GMT)

Presented By:

Josiah Coad

About the Speaker(s):

Josiah has interned at Facebook, Microsoft and the CIA. He is currently researching in reinforcement learning at Carnegie Mellon.

Join the Discussion:

Join our Discord community and talk with TD organizers and like minded DS/ML enthusiasts!

Join our Discord!