Building Similarity Search Engines Powered By Neural Networks

In this meetup, Jeremy will discuss how to leverage neural networks to power similarity search engines of arbitrary objects. Neural networks are great at learning representations of objects, and we'll walk through some of the practical challenges of using these vector representations to power a large scale search engine.

Jeremy works on the core machine learning team at Proofpoint, a leading cybersecurity company based in Silicon Valley.

Relevant topics:

• Neural codes for information retrieval
• Network architectures to learn good vector representations of objects
• Locally optimized product quantization for approximate nearest neighbors search

Schedule:
10:00am - Gather, meet other attendees, and discuss what you're currently working on
10:45am - Presentation (followed by Q&A)
12:00am - Study hall