Knowledge transfer from a source domain to a different but semantically related target domain has long been an important topic in the context of unsupervised domain adaptation (UDA) We propose a novel moment-based probability distribution metric termed dimensional weighted orderwise moment discrepancy (DWMD) for feature representation matching in the UDA scenario . Our metric function takes advantage of a series for high-order moment alignment, and we theoretically prove that our metric function is error-free . We further calculate the error bound of the empirical estimate of the DWMD metric in practical applications . Comprehensive experiments on benchmark datasets illustrate that our method yields state-of-the-art distribution metrics . We also calculate an empirical estimate the errorbound of the . DWMD metrics in practical use of the metric in the .DWMD metric function in practical purposes. We also describe how our method yield state-to-measure the accuracy of our method in our function. Comprehensive experiments to demonstrate that our . method yields the . method yield the state-for-the . dataset.

Links: PDF - Abstract

Code :


Keywords : metric - dwmd - method - domain - function -

Leave a Reply

Your email address will not be published. Required fields are marked *


Enjoy this blog? Please spread the word :)