Empirically, we show on a variety of datasets that the marginalized kernel mean estimator obtains much lower estimation error than the existing ...
確定! 回上一頁