This is a two-hour workshop over two days on the theme of Anomaly Detection in AI Applications. The first day is a lecture and demo on using the Anomaly Detection APIs for univariate and multivariate anomaly detection. The second day is a group discussion that will include discussions with the IBM Research AI Application teams.
In this lab-based tutorial, we present a web-based time series anomaly detection tool - a new scikit-learn compatible toolkit specialized for the time series based anomaly detection problem. The key focus of our tutorial includes the design and development of a anomaly detection pipeline, a zero-configuration interface for automated discovery of an anomaly pipeline for any given dataset (univariate and multi-variate), a set of 5 frequently used workflow empirically derived from past experiences, an scalable technique for conducting efficient pipeline execution.
Each API in the Anomaly Detection API addresses the challenges of enabling different type of anomaly detection. The APIs we will use in the workshop are:
- Univariate Anomaly Detection API
- Multivariate Anomaly Detection API
- Regression Based Anomaly Detection API
- Mixture-Model Based Anomaly Detection API
- Semi-supervised Anomaly Detection API
Prerequisites Background in Data Science or Computer Science
Take part in a day of interactive learning, as we explore that application of Distributed AI in real-world scenarios and share how the creative use of the Anomaly Detection APIs are being used to solve major problems.
TJThomas JackmanOffering Manager, Research AI Business Development Research, T&IPIBMDPDhaval PatelSTSMIBM
- Review the design and development of a anomaly detection pipeline, a zero-configuration interface for automated discovery of an anomaly pipeline for any given dataset (univariate and multi-variate)
- Distribution of knowledge check to practice application of APIs at home
- Modernize your approach by leveraging a state of the art Anomaly Detection APIs in our IBM Research JupyterLab