Registration: https://goo.gl/forms/b6lXh30Z64IePKGy2
RSVPs ARE NOT COMPLETE UNTIL YOU COMPLETE THE REGISTRATION AT THE LINK: Nametags will be printed in accordance with the google form you fill out :) ! Thanks!!
Abstract:
How do you get a machine learning system to deliver value from big data in a real world setting?
Turns out that 90% of the effort required for success in machine learning is not the algorithm or the model or the learning its the logistics. That 90% comes from many things, including the need to stage & deploy multiple versions of each model, to carefully collect & curate updated training data & to monitor model performance. Lately we have added scale, speed & the need to handle multiple machine learning frameworks at the same time to make the problem more difficult.
There is a way to make this easier & more effective the rendezvous architecture. This new design for model & data management is based on a streaming approach in a microservices style. It makes use of containerization & orchestration to solve many of the problems involved in continuous deployment of machine learning models. In presenting the rendezvous architecture, Ill cover techniques for effective model-to-model comparisons, for model deployment & management in production- including importance of new models on stand-by, & model monitoring. Finally, Ill talk about how a DataOps style of organization matches the flexibility offered by the rendezvous approach for machine learning.
Speaker:
Ted Dunning is Chief Applications Architect at MapR Technologies & a board member for the Apache Software Foundation, as well as PMC member/ committer of the Apache Mahout, Apache Zookeeper & Apache Drill projects & served mentor for several incubator projects. He has contributed to clustering, classification, matrix decomposition algorithms in Mahout & to the new Mahout Math library. He designed the t-digest algorithm used in several open source projects & by a variety of companies.
Ted was chief architect behind the MusicMatch (now Yahoo Music) & Veoh recommendation systems, built fraud-detection systems for ID Analytics (LifeLock) & has 24 issued patents to date. Ted has a PhD in computing science from University of Sheffield. He is on Twitter as @ted_dunning.
Agenda:
6:00 - 6:30 Registration & Networking
6:30 - 6:45 Introduction by The Hive
6:45 - 6:55 Introduction of speaker by Ellen Friedman
6:55 - 7:40 Talk by Ted Dunning
7:40 - 8:00 Q&A