Alex Combessie

Dataiku

Talk Abstract: The Making Of a Time Series Forecasting automated pipeline.

The story of my 20%-side-project at Dataiku: how I designed and created a visual plugin for business users and data scientist to forecast time series automatically without code. A journey from understanding the literature of time series forecasting, packaging the latest open source methods in a single pipeline, and making an interface for business users.

Bio: I have been a happy data scientist at Dataiku for two years, applying Machine Learning (but not only) to solve real issues. To me, putting a data project in production is like writing a good Haiku. Prior to that, I was involved in building the data science team at Capgemini Consulting. I started my career in economics, so I like my models interpretable as well as deep. I am into travelling, learning new things and crafting useful products.

 

Saturday April 13th , 2019
Auditorium
9:00 am-
5:00 pm
This free day will feature four-stream rooms with over 500 people attending. A day of lectures with over 40 top speakers. It will also feature our partners in our exhibitor section of the event. An entire day to learn, mingle and be inspired. For the people by the people! Please…
Sector: