Cross-Validation

Workshop: Cross-validation for feature selection in supervised machine learning models

Cross-validation is a technique that can prevent over-fitting of a model; it is used to generate a model that will perform best on an independent data-set.

Using a real-life example that EVERY student will be able to relate to, we will show how cross-validation works including a code demo.

Attendees will leave the workshop with an intuitive understanding of common machine learning terms such as supervised models, over-fitting, features, test vs training data, bias, and variance, as well as some examples of linear and logistic regression which they may have learned about in math class.

Code will be made available. No downloads are required; those who want to follow along should log into notebooks.azure.com with their MyPortal ID and password. (Use your 8-digit number followed by "@fhda.edu" for redirect to MyPortal sign on)