Ordinal multivariate regression with regularization (Lasso and elastic net)

6 views (last 30 days)
I'm looking to do a regression of 80 independent variables (continuous values, though many highly correlated) against 750 responses which are ordinal (1,2,3,4). The lassoglm function will perform a linear model regression with regularization for a binomial outcome but not a multinomial one. mnrfit will do the multinomial (ordinal) regression but without the regularization. Does anyone know how I can use regularization with the multinomial regression?
And a related question, does anyone know how to obtain p-values for optimal predictor coefficients with lasso regularization?
Many thanks in advance.

Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!