That’s because for data scientists, our annual user conference is always brimming with the kind of breakout sessions that can help them to do their jobs better.
This year is no exception. Once again, there will be plenty of illuminating sessions for data scientists to choose from. All are worth checking out, of course. But for the sake of brevity here, I’ve outlined three.
When Real Data Doesn’t Match Science Theory
(March 19, 2:15 p.m.)
Titan America LLC has been using Domo to build predictive models that have surpassed the state-of-the-art in accuracy and explanatory power.
But sometimes even the best data science models can fall short of expectations in a data science production environment.
Learn how one of the country’s premier cement and building materials producers is now using Domo to refine production data science models to drive actionable insights that improve their products.
Titan America LLC’s David Brader and Noah Erickson will discuss the scientific process they use to develop models—including how they identify mediating and moderating relationships in their data that would otherwise generate misleading results.
They will also demo their data science production pipeline, featuring data sanitation alerts, multiple model comparisons, end user dashboards, and next-step prescriptive actions to improve cement production.
Partnering With Domo for Data Science Success
(March 18, 2 p.m.)
In this session, Asure Software shares its experience in getting started with data science.
The maker of a human capital management (HCM) solution offers both technical and business insights into how to get off on the right foot when starting your data science journey.
For advanced data scientists, Asure’s Bruce Harris and Ulises Gonzalez-Guerra demonstrate how they leverage Domo’s Data Science Suite to deploy custom scripts that create comprehensive data profiles in only a few minutes.
As such, data profiling and cleaning—a process that often requires a substantial time commitment—can be done quickly, automatically, and in a way that supports your data science production pipeline.
Finally, Bruce and Ulises show the audience how they calculate ROI to determine their data science investments and priorities.
Data Wrangling Names and Locations with Scripting Tiles or the R Plug-In
(March 18, 3 p.m.)
Randstad is a Dutch multinational human resource consulting firm that focuses on developing a thorough data science pipeline prior to estimating predictive models.
The company’s pipeline considers and employs best practices for data structuring and feature engineering as they relate to their use cases.
Randstad’s demo highlights how to use Domo’s Data Science Suite for structuring categorical data, using fuzzy matching for merging datasets, and novel ways to represent location data within a data science modeling framework.