Abstract:
|
For massive data with super-large sample size, it is computationally infeasible to obtain maximum likelihood estimates for unknown parameters, especially when the estimators do not have closed-form solutions. In this talk, I will present fast leveraging algorithms to efficiently approximate the maximum likelihood estimates in logistic regression models with binary responses, one of the most commonly used models in practice for classification. I will also present some theoretical results on consistency and asymptotic normality of the estimators. Synthetic and real data sets are used to evaluate the practical performance of the proposed methods.
|