The basic concepts of validation, hyper-parameters, over-fitting and regularization are introduced in the context of polynomial regression.
When and Where?
This workshop runs multiple times.
10AM IST, Saturday Feb 26th, 2022 Register
10AM IST, Sunday Mar 13th, 2022 Register
10AM IST, Saturday Mar 26th, 2022 Register
10AM IST, Saturday Apr 09th, 2022 Register
Details and Outcomes
- bias and variance, underfitting and overfitting
- training, testing, and validation sets, including cross-validation for hyper-parameters
- regularization as a different approach to complexity control
At the end of this workshop, you will understand the basic concepts of machine learning, and be able to identify them in any new model you see.
Dr. Rahul Dave
Rahul is co-founder of Univ.AI. He was previously a lecturer at Harvard University. He was on the original team for Harvard’s famous Data Science course, cs109, and has taught machine learning, statistics, and AI courses, both at Harvard and at multiple conferences and workshops. Some of his more popular offerings have been the Data Scientist Training for Librarians workshops in Boston and Copenhagen, Machine Learning for Suits at the Open Data Science Conference, continuing versions of cs109, and the am207 course on Bayesian Statistics and Generative Models at Harvard. Rahul is an accomplished computational scientist with a strong programming background and a veteran cosmologist. His Ph.D. thesis in cosmology and astrophysics at the University of Pennsylvania involved both high performance computing and bayesian statistics, and was one of the first works introducing dark energy. His subsequent work in Solar System astronomy and large scale astronomy databases at the ADS took him in the direction of machine learning and AI. Rahul is passionate about teaching, and a big believer in exposing big, ‘researchy’ ideas early on, to students. Follow him at @rahuldave on Twitter.