Answers are grounded in this paper and cite their sources.
The article considers the multi-step ADALINE training algorithm when using the correntropy information criterion as a learning criterion, determines the conditions for the convergence of the algorithm, and shows that in the steady state the resulting estimate is unbiased. The importance of choosing the width of the Gaussian core, which affects the convergence rate of the estimation algorithms and the error in the steady state, is noted, and the feasibility of developing procedures for adaptive correction of the core width is indicated.
No takes yet. Share an insight, caveat, or question.
Rudenko et al. (2020) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: