Bay Area Women in Machine Learning & Data Science

Our next meetup will be a series of presentations on hyperparameter optimization and how to use various software packages to find a set of optimal hyperparameters for your machine learning model. Model selection via hyperparameter optimization is an important part of machine learning and we will discuss both the very basic and sophisticated methods for tuning models. Speaker: Erin Craig Title: Hyperparameter Optimization: Grid Search and Bayesian Optimization Abstract: When building a model, how do you select its hyperparameters? Grid search and bayesian optimization are two common methods for hyperparameter optimization; each with its own set of strengths and drawbacks. We begin this talk with a brief overview of these two methods, and then look at a case study to compare results of manual tuning, grid search and Bayesian optimization when predicting 30-day readmission from electronic health records.