Aadyot Bhatnagar on building an end-to-end machine learning framework for time series data.
Aadyot Bhatnagar, is a Senior Research Engineer at Salesforce, and co-creator of Merlion an open source framework for applying machine learning on time series data. Merlion supports a wide range of time series learning tasks including forecasting, anomaly detection, and change point detection. Equally important, Merlion is an end-to-end framework that covers loading and transforming data, building and training models, post-processing model outputs, and evaluating model performance.
Highlights in the video version:
Merlion and basic components of a time series library
User persona, specific features, accuracy of models in APM world
Research to production
Speed, scale, and roadblocks in time series
Distributed computing and synthetic data
Will there be pre-trained models for time series?
Community contributors, use cases, and multivariate time series challenges
Roadmap for the next six to twelve months
What is needed to install Merlion?
- A video version of this conversation is available on our YouTube channel.
- Reza Hosseini and Albert Chen: Building a flexible, intuitive, and fast forecasting library
- Sean Taylor: Changes to the data science role and to data science tools
- Sercan Arik: Neural Models for Tabular Data
- Murat Özbayoğlu: Financial Time Series Forecasting with Deep Learning
- Machine Learning Trends You Need To Know
- The Graph Intelligence Stack
- fastdup: Introducing a new free tool for curating image datasets at scale
- Previous podcast episodes pertaining to time series and forecasting.
[Image: from Unravel sourced via Infogram.]