medium10 pts · first answer countsstatisticsregressionregularisation
Ridge versus lasso
You fit a regression with 200 candidate predictors and want a model that keeps only a few. Which regularisation tends to set some coefficients exactly to zero?
Discussion 0
Join free to post in the discussion.
Related questions
- What the central limit theorem saysData Scientist
- Two independent servers both upData Scientist
- Churn rate among mobile usersData Scientist
- Expected value of a raffle ticketData Scientist
- Converting units and the SDData Scientist
- Expected sign-ups from 100 visitsData Scientist
- Arrivals in ten minutesData Scientist
- Beyond 1.96 standard deviationsData Scientist