Abhishek Murthy
Abhishek Murthy, Ph.D., is a Senior Principal ML/AI Architect at Schneider Electric (SE) in Boston, Massachusetts.
At SE, Dr. Murthy develops Machine Learning (ML) algorithms on sensor data, which are critical for the company's energy technology offers. He is currently focused on leveraging these technologies to improve SE's service offerings for data centers.
Dr. Murthy is also an Adjunct Faculty Member at Northeastern University, where he teaches machine learning algorithms for the Internet of Things (IoT).
He holds a Ph.D. in Computer Science from Stony Brook University, State University of New York, and an M.S. in Computer Science from the University at Buffalo. His doctoral research, supported by a National Science Foundation (NSF) Expedition in Computing, focused on developing algorithms for automatically establishing the input-to-output stability of dynamical systems.
Prior to joining SE, he led the Data Science Algorithms team at WHOOP and served as a Senior Data Scientist at Signify (formerly Philips Lighting), where he led research on IoT applications for smart buildings. Dr. Murthy is an active contributor to the field, with several publications, more than 210 citations, 21 awarded patents, and over 40 pending applications.
Session
The time series machine learning community has begun adopting foundational models for forecasting and anomaly detection. These models, such as TimeGPT, MOMENT, Morai, and Chronos, offer zero-shot learning and promise to accelerate the development of AI use cases.
In this talk, we'll explore two popular foundational models, TimeGPT and MOMENT, for Time Series Anomaly Detection (TSAD). We'll specifically focus on the Novelty Detection flavor of TSAD, where we only have access to nominal (normal) data and the goal is to detect deviations from this norm.
TimeGPT and MOMENT take fundamentally different approaches to novelty detection.
• TimeGPT uses a forecasting-based method, tracking observed data against its forecasted confidence intervals. An anomaly is flagged when an observation falls sufficiently outside these intervals.
• MOMENT, an open-source model, uses a reconstruction-based approach. The model first encodes nominal data, then characterizes the reconstruction errors. During inference, it compares the test data's reconstruction error to these characterized values to identify anomalies.
We'll detail these approaches using the UCR anomaly detection dataset. The talk will highlight potential pitfalls when using these models and compare them with traditional TSAD algorithms.
This talk is geared toward data scientists interested in the nuances of applying foundational models for TSAD. No prior knowledge of time series anomaly detection or foundational models is required.