My Software

Sigopt Open Source

Adaptive experimentation is most commonly used in practice for machine learning hyperparameter optimization, in which one seeks to improve say, the accuracy of a model by varying its underlying hyperparameters.

I contribute to SigOpt Open Source, It’s a package for adaptive experimentation and sequential optimization offering both hosted and in-memory solutions. My team has taken great pains to make sure it’s performant, fast, and above all else, reliable.