Put AI to work
June 26-27, 2017: Training
June 27-29, 2017: Tutorials & Conference
New York, NY

AI for structured business data

Vikash Mansinghka (MIT), Richard Tibbetts (Empirical Systems)
1:30pm5:00pm Tuesday, June 27, 2017
Impact of AI on business and society
Location: Sutton South/Regent Parlor Level: Beginner
Secondary topics:  Machine Learning
Average rating: ****.
(4.50, 2 ratings)

Prerequisite Knowledge

  • Experience with structured data in a business setting
  • Familiarity with decision support or business intelligence
  • A working knowledge of SQL or Python (i.e., the ability to read short snippets of code)

What you'll learn

  • Learn how AI can help business analysts and data scientists analyze structured business data and provide decision support
  • Explore new AI technologies, such as probabilistic programming, deep learning, and Bayesian nonparametrics, that integrate well with software tools for standard data science workflows, including both open source and commercial products, guide business analysis to statistically meaningful questions from structured business data, and help junior data scientists answer them in minutes

Description

1:30pm – 3:00pm—Vikash Mansinghka: Open source AI software for structured business data

Vikash Mansinghka offers a brief overview of AI and data science strategy for a broad business and technical audience, followed by a hands-on tutorial on open source AI software for structured business data. Examples will be drawn from recent state-of-the-art probabilistic AI systems, including (but not limited to) the open source probabilistic computing stack produced by the MIT Probabilistic Computing Project. Hands-on exercises will focus on probabilistic data search and cleaning but will include the option to work from templates for virtual experiments and time series structure discovery, depending on interest.

Topics include:

  • Probabilistic data search: Finding business data that is predictively relevant for a variable or record of interest. Applications include fraud and audit, document and vendor discovery, and exploratory data analysis.
  • Probabilistic data cleaning: Automatically identifying likely multivariate anomalies and interactively narrowing down on potential causes. Applications include identifying over- and underperformers and fixing ETL errors.
  • Simulating virtual experiments to answer “What if?” questions when the real business experiment would be too costly.
  • Discovering structure in time series data, such as seasonal trends and short-term deviations from long-term increases or declines.


3:30pm – 5:00pm—Richard Tibbetts: AI for structured business data

Businesses have spent decades trying to make better decisions by analyzing structured data. New AI technologies like probabilistic programming and Bayesian nonparametrics are just beginning to transform this process. Richard Tibbetts explores AI that guides business analysts to ask statistically sensible questions and lets business professionals and junior data scientists answer in minutes questions that previously took hours for trained statisticians. In this workshop, you’ll learn how to use AI to answer questions about the probable implications of messy structured data via a combination of open source and commercial tools. This tutorial focuses on how to go beyond understanding your data to act based on data insights.

Photo of Vikash Mansinghka

Vikash Mansinghka

MIT

Vikash Mansinghka is a research scientist at MIT, where he leads the Probabilistic Computing Project, and a cofounder of Empirical Systems, a new venture-backed AI startup aimed at improving the credibility and transparency of statistical inference. Previously, Vikash cofounded a venture-backed startup based on his research that was acquired by Salesforce, was an advisor to Google DeepMind, and held graduate fellowships at the National Science Foundation and MIT’s Lincoln Laboratory. He served on DARPA’s Information Science and Technology advisory board from 2010 to 2012 and currently serves on the editorial boards for the Journal of Machine Learning Research and Statistics and Computation. Vikash holds a PhD in computation, an MEng in computer science, and BS degrees in mathematics and computer science, all from MIT. His PhD dissertation on natively probabilistic computation won the MIT George M. Sprowls dissertation award in computer science, and his research on the Picture probabilistic programming language won an award at CVPR.

Photo of Richard Tibbetts

Richard Tibbetts

Empirical Systems

Richard Tibbetts is CEO of Empirical Systems, an MIT spinout building an AI-based data platform for organizations that use structured data to provide decision support. Previously, he was was founder and CTO at StreamBase, a CEP company that merged with TIBCO in 2013, as well as a visiting scientist at the Probabilistic Computing Project at MIT.