Nathaniel Cook - Forecasting Time Series Data at scale with the TICK stack

Thanks! Share it with your friends!

You disliked this video. Thanks for the feedback!

Added by admin
28 Views
Description
Forecasting time series data across a variety of different time series comes with many challenges. Using the TICK stack we demonstrate a workflow that helps to overcome those challenges. Specifically we take a look at the Facebook Prophet procedure for forecasting business time series.

Abstract
Forecasting time series data can require a significant amount of attention in order to get reliable results. As the number and variety of time series increases it becomes too expensive to manage each forecast individually. Using the TICK stack we demonstrate a workflow that helps to reduce the amount of attention each forecast needs. This is accomplished by using the procedure called Prophet, which was recently open sourced by Facebook.

The basic idea of this procedure is two fold:

Reduce the amount of effort to train and maintain a single forecast.
Automatically surface the forecasts that are performing poorly.
To reduce the amount of effort per forecast, the Facebook Prophet algorithm is a simple model with a few well understood parameters. By automatically surfacing forecasts that perform poorly, effort need only be spent when specific forecasts need attention.

The last piece needed to make this process scale is a single set of tools that follow the workflow. We demonstrate how the TICK stack can be used to manage forecasting time series at scale, using InfluxDB to store the data and Kapacitor to manage and surface forecasts.

www.pydata.org

PyData is an educational program of NumFOCUS, a 501(c)3 non-profit organization in the United States. PyData provides a forum for the international community of users and developers of data analysis tools to share ideas and learn from each other. The global PyData network promotes discussion of best practices, new approaches, and emerging technologies for data management, processing, analytics, and visualization. PyData communities approach data science using many languages, including (but not limited to) Python, Julia, and R.

PyData conferences aim to be accessible and community-driven, with novice to advanced level presentations. PyData tutorials and talks bring attendees the latest project features along with cutting-edge use cases.
Category
PyData Seatle 2017

Post your comment