Feature Extraction

It's all in the feature extraction! Understanding how the latest machine learning algorithm works is cool, but once you start working with data, you find out that there are other things equally as important, or maybe even more so. How you turn your raw data into features used in learning and modelling methods can make a huge difference in their predictive capability. So how do we extract these features? What kind of transformations should we apply? When should you normalize?

We will be using the rich open government data resource from data.gov (http://data.gov/) https://www.data.gov/open-gov/ to demonstrate how features can be defined and extracted from real data and what to do with them once you define them.

Base concepts micto topic - Linear Modelling!

How do you combine multiple predictor variables in one model and why combine them linearly? We will wrap up the meet up with a little did bit of statistics to expand and solidify the basics of data science.