Presented By
O’Reilly + Cloudera
Make Data Work
March 25-28, 2019
San Francisco, CA

Modern techniques for building robust deep networks

Sricharan Kumar (Intuit )
11:50am12:30pm Thursday, March 28, 2019
Average rating: ****.
(4.29, 7 ratings)

Who is this presentation for?

  • Data scientists at all levels and students

Level

Intermediate

Prerequisite knowledge

  • Basic knowledge of machine learning

What you'll learn

  • Learn best practices for deploying robust machine learning models.

Description

Machine learning is delivering immense value across industries. However, machine learning models can sometimes produce overconfident results, with the potential for unintended—even catastrophic—outcomes.

Kumar Sricharan draws upon his experience in the aerospace, consumer, finance, and healthcare industries to explain why existing machine learning techniques work well in certain situations, but can fail in others; describe how today’s data scientists typically approach the problem; and share recent innovative techniques that enhance robustness of models, including deep neural networks.

Photo of Sricharan Kumar

Sricharan Kumar

Intuit

Kumar Sricharan is a principal data scientist responsible for leading Intuit’s Machine Learning (ML) Research Group. His team focuses on cutting-edge ML problems and applications for financial artificial intelligence, combining financial domain knowledge with information in data to produce more accurate and explainable systems. Projects include extraction of information from financial documents, chatbots for financial conversations, mining tax forms to enable compliance, understanding transactions to power financial advice, and financial forecasting. Previously, he spent five years at Xerox PARC as a senior research scientist and program manager for self-learning systems. During his time at PARC, Kumar led a team focused on building learning algorithms that can exploit rich feedback in conjunction with unlabeled data to overcome the need for prohibitively large numbers of labeled examples and make these systems explainable. His other research interests include anomaly detection for large, unstructured data and nonparametric large sample estimation, and his research has resulted in nearly 30 papers in refereed conferences and journals and several accompanying patents. He’s an active member of the academic community and has served as a reviewer for multiple conferences and journals. He has also participated in several DARPA research projects, including the ADAMS program for detecting insider threat and XAI for building explainable AI agents. Kumar holds a PhD in electrical engineering from the University of Michigan.

Comments on this page are now closed.

Comments

Picture of Shradha Agrawal
Shradha Agrawal | DATA SCIENTIST
04/03/2019 3:38am PDT

Nice talk!! was wondering when the slide deck would be made available