Transfer learning for nonparametric classification: minimax rate and adaptive classifier (Q2656589)

The main contribution is the rate-optimal two-sample weighted K-NN classifier. The convergence analysis is fulfilled. It is to be noted that the proposed model automatically adapts to the unknown model parameters. The ideas of \textit{O. L. Lepski} et al. [Ann. Stat. 25, No. 6, 2512--2546 (1997; Zbl 0894.62041)], developed for nonparametric regression, were employed. The efficiency is demonstrated on three important examples: crowdsourcing, concept drift and data corruption.