Synopsis: The traditional Least Square estimation, KNN face severe overfitting issues when the dataset has high-dimensional features. Modern data mining regression techniques such as lasso and classification techniques such as SVM give a better estimation result in such a situation.
Start Time: Friday, November 9, 2018 1:00 PM
End Time: Friday, November 9, 2018 3:00 PM
Location: Alexander Library
Address: 169 College Avenue
Campus: College Avenue
Room: 415
City, State, Country: New Brunswick, NJ US
Fee: N/A
Sponsor: Rutgers University Libraries
Category: Workshop, Course
Contact Name: Ryan Womack
Contact Email: rwomack@rutgers.edu
Contact Phone: (848) 932-6107
Additional Information: The workshop intends to show how lasso and SVM work in Python. Compare the estimation result of Lasso with least square estimation, SVM with KNN in the high-dimensional setting.
Target Audience: Current Students,  Faculty,  Staff