Empirical Risk
A standard performance measure in machine learning that approximates the true expected risk using the average loss over a training set. It is the central concept in the Empirical Risk Minimization (ERM) framework, and its convergence properties depend on the complexity of the hypothesis class.
Timeline (last 90 days)
The core logic of SEED-MAX is to recursively compute the empirical risk, selecting at each step the candidate feature that can reduce the excess risk to the greatest extent to add to the set, maintaining a set of selected features and obtaining theoretical guarantees through a greedy recursive strategy.
The paper identifies the convergence behavior of empirical risk across different regimes: some regimes maintain an O(1/√N) rate, some regimes converge noticeably slower, and in some regimes the empirical risk no longer exhibits consistency.
All Facts (2)
The core logic of SEED-MAX is to recursively compute the empirical risk, selecting at each step the candidate feature that can reduce the excess risk to the greatest extent to add to the set, maintaining a set of selected features and obtaining theoretical guarantees through a greedy recursive strategy.
50%UnverifiedThe paper identifies the convergence behavior of empirical risk across different regimes: some regimes maintain an O(1/√N) rate, some regimes converge noticeably slower, and in some regimes the empirical risk no longer exhibits consistency.
50%