About the event
As part of the MagIC Seminar Series, NOVA IMS will welcome Filiz Karadağ, from the Department of Statistics at Ege University in İzmir, Türkiye, for the seminar Robust Estimation Methods in Ridge Regression with R Packages.
Organized by the NOVA Applied Economics & Analytics Lab, in collaboration with MagIC, the session will focus on the implementation of regularized regression and robust parameter estimation methods, with practical demonstrations using the R packages S-type.est, Styperidge.reg, and ridgregextra.
The presentation will explore how these methods can be used to reduce the impact of outliers and address multicollinearity. Topics will include VIF-guided automatic ridge parameter selection and the diagnostic tools available in ridgregextra, as well as the combination of robust estimation and ridge regularization through Styperidge.reg.
The methods will be demonstrated across datasets of different sizes, ranging from benchmark datasets to large-scale energy consumption data, while also addressing computational efficiency and scalability.
The session will also introduce the three R packages presented, explain their main features and practical applications, and provide a brief overview of the methodological research underlying their development.
