SigOpt collaborated with MLconf on a webinar to discuss best practices for metrics, training, and hyperparameter optimization for developing high-performing models. During this talk, SigOpt Head of Product Fay Kallel and Head of Engineering Jim Blomo shared insights that they’ve developed from working with a wide range of modelers on the SigOpt platform.
Watch this video to follow an applied fraud detection use case to learn how to:
- Discover, track, and compare many metrics to understand model behavior and connect machine learning to business outcomes
- Track, visualize, and analyze training runs to create baselines, interpret model convergence, and explore the modeling problem space
- Automate, accelerate, and scale hyperparameter optimization to identify configurations of your model that meet specific business needs