Dynamic replication algorithms aim at allocating, migrating and deleting copies of an object over various Internet hosts, according to the access patterns exhibited online, so as ...
This paper presents a corpus-based algorithm capable of inducing inflectional morphological analyses of both regular and highly irregular forms (such as broughtbring) from distrib...
Corpus-based methods for natural language processing often use supervised training, requiring expensive manual annotation of training corpora. This paper investigates methods for ...
Background: The development of effective frameworks that permit an accurate diagnosis of tumors, especially in their early stages, remains a grand challenge in the field of bioinf...
Andreas Keller, Nicole Ludwig, Nicole Comtesse, An...
Traditional classification methods assume that the training and the test data arise from the same underlying distribution. However, in several adversarial settings, the test set is...