Back To Schedule
Saturday, October 20 • 2:00pm - 2:30pm
The Future of Prediction: Utilizing Unsupervised Learning

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Advances in Artificial Intelligence over the past few years have focused on the ability to build better predictive models using a variety of techniques, not the least of which is deep learning. These approaches, however, require massive datasets and restrict their utility for enterprise grade problems. Complicating matters further, the vast majority of enterprise data is unlabeled creating an additional class of challenges.

The future of prediction lies upstream in the analytical process – in the unsupervised learning techniques of segmentation, anomaly detection and hotspot detection. Here, unencumbered by labels, or even the requirement to know what you are looking for, lies the keys to improving the performance of prediction. This nascent form of machine learning has already had enormous impact on some of today’s largest enterprises across finance, government and healthcare.

In this session, Gunnar Carlsson, President & Co-Founder at Ayasdi, will dive deeper into unsupervised learning and why this is the next frontier in AI. He'll break the problem down technically and practically – discussing feature generation, dimensionality reduction and how to get started. In doing so she will explore key enterprise uses cases ranging from fraud detection, financial crime and population health

avatar for Neha Agarwal

Neha Agarwal

Director of Engineering, Thumbtack
Scott Castle is GM of the Data Business at Sisense and served as VP of Product for Periscope Data prior to its merger with Sisense in 2019. At Periscope Data, Scott oversaw product strategy, planning, design, and delivery. He brings over 20 years of experience in software development... Read More →

Saturday October 20, 2018 2:00pm - 2:30pm
Ballroom # 406