Artificial Intelligence (AI) In Real Life

Abstract: The greatest challenge when building a high-performance model isn't about choosing the right algorithm or doing hyperparameter tuning: it is about getting high quality labeled training data. Without good data, no algorithm, even the most sophisticated one, will deliver the results needed for real-life applications. And with most modern algorithms (such as Deep Learning models) requiring huge amounts of data to train, things aren't going to get better any time soon.

Active Learning is one of the possible solutions to this dilemma, but quite surprisingly, left out of most data science conferences and computer science curricula. By the end of this presentation, you will learn the importance of Active Learning and its application to make AI work in the real world.

Bio: Jennifer Prendki is the founder and CEO of Alectio. The company is the direct product of her beliefs that good models can only be built with good data, and that the brute force approach that consists in blindly using ever larger training sets is the reason why the barrier to entry into AI is so high. Prior to starting Alectio, Jennifer was the VP of Machine Learning at Figure Eight, the company that pioneered data labeling. She has been Chief Data Scientist at Atlassian and Senior Manager of Data Science in the Search team at Walmart Labs. Jennifer has spent most of her career creating data-driven cultures, succeeding in sometimes highly skeptical environments. She is particularly skilled at building and scaling high-performance machine learning teams and is known for enjoying a good challenge. Trained as a particle physicist (she holds a PhD in particle physics from Sorbonne University), she likes to use her analytical mind not only when building complex models but also as part of her leadership philosophy. She is pragmatic yet detail-oriented. Jennifer also takes great pleasure in addressing both technical and nontechnical audiences alike at conferences and seminars and is passionate about attracting more women to careers in STEM.

Abstract: There’s never been a more exciting time to work in networking and networked applications.  This talk will examine how machine learning (ML) benefits networking by focusing on four examples.  First, we’ll examine for Intent-Based Networking (a modern architecture for designing and operating a network) and how ML can be used to increase visibility, diagnose problems and identify associated remedies, and provide assurance that the network is operating as intended.  Next, we’ll look at how the move from today’s Cloud-based ML to the promising approach of Distributed ML across Edge and Cloud can lead to improved scalability, reduced latency, and improved privacy.  We’ll also discuss how to identify what devices are on the network and how the network should treat those devices.  Lastly, in the context of ever-growing security threats, we examine how ML can be applied to address the challenge of malware sneaking in an encrypted flow.  Specifically, how we can detect malware hidden in encrypted flows without requiring decryption of those flows.  It is noteworthy that while ML is often associated with reducing privacy, this example showcases how an elegant application of ML can both preserve privacy and reduce complexity.

Bio: John is VP/CTO of Cisco's Enterprise Networking Business (Cisco's largest business) where he drives the technology and architectural direction in strategic areas for the business. This covers the broad Cisco portfolio including Intent-Based Networking (IBN), Internet of Things (IoT), wireless (ranging from Wi-Fi to emerging 5G), application-aware networking, multimedia networking, indoor-location-based services, connected vehicles, machine learning and AI applied to the aforementioned areas, and deep learning for visual analytics.

Previously, John was Lab Director for the Mobile & Immersive Experience Lab at HP Labs. The MIX Lab conducted research on novel mobile devices and sensing, mobile client/cloud multimedia computing, immersive environments, video & audio signal processing, computer vision & graphics, multimedia networking, glasses-free 3D, next-generation plastic displays, wireless, and user experience design.

John received a number of honors and awards including IEEE Fellow, IEEE SPS Distinguished Lecturer, named “one of the world’s top 100 young (under 35) innovators in science and technology” (TR100) by MIT Technology Review, received a Certificate of Honor for contributing to the US Digital TV Standard (Engineering Emmy Award 1997) and his work on media transcoding in the middle of a network while preserving end-to-end security (secure transcoding) was adopted in the JPSEC standard. He has published over 100 papers, including receiving 5 best paper awards, and has about 75 granted US patents. John also has strong collaborations with the academic community and was a Consulting Associate Professor of EE at Stanford (2000-09), and frequently lectures at MIT.  He received his B.S., M.S., and Ph.D. in EECS from MIT.