Presented by Arshak Navruzyan, VP of Product at Argyle Data.
Discount for Meetup Members: Get 30% off with code VW30 at http://www.zipfianacademy.com/workshops/introduction-machine-learning-vowpal-wabbit
Date and Time: September 13 & 14 from 9-4:30pm (2 full days)
Summary: Small-group training in using Vowpal Wabbit. Access to workstations and data sets are included.
Level: Designed for ML beginners. Analysts and MBAs welcome.
Vowpal Wabbit (VW) is one of the hidden gems of machine learning. It is super-fast (can learn 4 million features per second per node), scalable and extremely flexible.
In this workshop we will get hands on with VW and build terafeature models. You will learn how to build Internet-scale machine learning models from data sets that are both tall (millions/billions of examples) and wide (many thousands of features). We will cover everything you need to know to put ML models into a production application including:
- basics of online
- choice of optimizers
- loss functions
- feature factorization and hashing
- importance weights
- holdouts vs. progressive validation
- distributed learning through allreduce
- model evaluation and management
- applying models / predictions
Since no coding is required, this workshop is perfect for machine learning beginners. MBAs, business/finance analysts, business intelligence professionals and product managers are welcomed to attend.
Open source VW software will be used during the workshop. Please bring a (decent) laptop.
About Arshak Navruzyan
Arshak's objective is to make distributed systems and machine learning accessible to any organization or individual that wants to transform the world through data.
He is currently VP of Product Management at Argyle Data focused on petabyte-scale risk management applications using machine learning and Hadoop. Previously he held senior engineering and product management roles at Alpine Data Labs, Endeca and Oracle.
He is a contributor to the Apache Accumulo project and the organizer of San Francisco Machine Learning Meetup group.