Predictive Modeling and the Data Pipeline that Feeds It - Jeff Friesen and Amber Rivera
plus Pizza and soda!
Are you data-science curious? Long-time developer but have never done your own machine learning project start to finish? Or maybe you have, are now addicted, and looking for more ideas. This talk is for you. Jeff Friesen and Amber Rivera recently teamed up to see whether it’s possible to predict a household’s likelihood to install an energy efficiency upgrade. This type of “propensity modeling” is used across industries to determine a consumer’s likelihood to take an action. Jeff will describe his journey collecting and transforming public and private datasets in the building sector. Starting with a pile of scripts and instructions on which order to run them, he upgraded to a mix of tools and scripts. Now everything is being rebuilt using Google Cloud Storage, Big Query, Dataprep and custom Clojure and Python code. Amber will walk through one application of that data - applying supervised machine learning techniques to build a classifier for prediction, based on historical data. She’ll also explain why a business would care about doing this, by going into the cost and benefit tradeoffs of targeting potential customers based on this technique.
Jeff Friesen is CTO and Co-Founder of Radiant Labs, a software company that models the energy and financial opportunities to promote energy efficiency, solar and electric vehicle upgrades. Using large datasets, Radiant Labs is modeling entire cities and counties at a time.
Amber Rivera is a data science consultant for energy, media and government clients, and a graduate of the data science immersive program at Galvanize in Denver. Before diving into machine learning, Amber used data analytics to drive increased audience engagement for the public media project “Inside Energy”, and worked at a solar financing startup modeling risk in the company’s solar asset portfolio.
Food and refreshments are kindly provided by the good folks at KJ Technical. http://kjtechnical.com/
We will have a selection of pizza, including vegetarian and GF options.
Note: try to arrive a little before 6pm. The entrance doors to the building are locked at 6pm. We will leave telephone contact info posted on the door so you can get in after 6.