Deep learning for anomaly detection
Who is this presentation for?Data scientists or analysts
In many business use cases, it’s frequently desirable to automatically identify and respond to situations where data deviates from expected notions of normal (anomalies). This may be in the context of flagging anomalous samples to be removed from a dataset, identifying malicious network traffic data that needs to be blocked, abnormal financial transactions that may be indicative of fraud, abnormal images in medical imaging data, system health monitoring, and such. In many of these cases, there may be no clear definition of what represents an anomaly, necessitating the use of unsupervised techniques. Furthermore, when working with large, high-dimensional, multivariate data, traditional approaches for anomaly detection are no longer suitable.
Nisha Muktewar and Victor Dibia explore a set of deep learning algorithms for anomaly detection including sequence to sequence models (Seq2Seq), variational autoencoders (VAEs), and generative adversarial networks (GANs) based approaches. In addition, they provide practical guidance for the successful implementation of anomaly detection systems within enterprises across key metrics like interpretability, reduction of false positives, and scalability. You’ll see benchmark results from applying them using several datasets and learn industry-best practices for deploying each model. They provide instruction for how and why these algorithms work through a demo with a working prototype.
- A basic understanding of deep neural networks
What you'll learn
- Learn why and when to use deep learning for anomaly detection
- See in-depth coverage of various deep learning model approaches and how they can be applied for anomaly detection, performance benchmarks for various deep learning models for a given dataset, and example use cases, product possibilities, and best practices
Cloudera Fast Forward Labs
Nisha Muktewar is a research engineer at Cloudera Fast Forward Labs, which is an applied machine intelligence research and advising group part of Cloudera. She works with organizations to help build data science solutions and spends time researching new tools, techniques, and libraries in this space. Previously, she was a manager on Deloitte’s actuarial, advanced analytics, and modeling practice leading teams in designing, building, and implementing predictive modeling solutions for pricing, consumer behavior, marketing mix, and customer segmentation use cases for insurance and retail and consumer businesses.
Cloudera Fast Forward Labs
Victor Dibia is a research engineer at Cloudera’s Fast Forward Labs, where his work focuses on prototyping state-of-the-art machine learning algorithms and advising clients. He’s passionate about community work and serves as a Google Developer Expert in machine learning. Previously, he was a research staff member at the IBM TJ Watson Research Center. His research interests are at the intersection of human-computer interaction, computational social science, and applied AI. He’s a senior member of IEEE and has published research papers at conferences such as AAAI Conference on Artificial Intelligence and ACM Conference on Human Factors in Computing Systems. His work has been featured in outlets such as the Wall Street Journal and VentureBeat. He holds an MS from Carnegie Mellon University and a PhD from City University of Hong Kong.
Leave a Comment or Question
Help us make this conference the best it can be for you. Have questions you'd like this speaker to address? Suggestions for issues that deserve extra attention? Feedback that you'd like to share with the speaker and other attendees?
Join the conversation here (requires login)
Premier Diamond Sponsors
Premier Exhibitor Plus
For conference registration information and customer service
For more information on community discounts and trade opportunities with O’Reilly conferences
For information on exhibiting or sponsoring a conference
For media/analyst press inquires