It's Time for Py: Time Series Clustering and Anomaly Detection in Python
Thunder Talk
Time series data provides helpful insights about trends, seasonality and variance in applications varying from finance to the environment to personal sensors. But what do you do when you have hundreds or thousands of time series?
Enter time series clustering. In this talk, Susan Devitt, Senior Data Scientist, will introduce two algorithms for time series clustering - Shape Based Distance and Dynamical Time Warp - implemented in the python package dtwclust. Taking a real world example using daily financial transaction data, she will show you when to consider clustering and how it can be used for anomaly detection.
Presented by
Susan Cameron Devitt
Susan has over a decade of experience in the full life cycle of data science projects from data engineering and quality assurance to model deployment, interpretation, and data visualization in a variety of industries including biotechnology, retail and oil and gas. Susan has a PhD from the University of California and completed a Postdoctoral fellowship at Harvard University. Before joining Sense Corp, Susan taught data analysis and statistics at the University of Texas at Austin and the University of Florida. She is a leading member in the Sense Corp Data Science and Engineering Center of Excellence where she drives innovation and provides internal and external training and development in data science.