Recently, spammers have proliferated "image spam", emails which contain the text of the spam message in a human readable image instead of the message body, making detect...
This paper investigates reinforcement learning (RL) in XCS. First, it formally shows that XCS implements a method of generalized RL based on linear approximators, in which the usu...
Using a specific machine learning technique, this paper proposes a way to identify suspicious statements during debugging. The technique is based on principles similar to Tarantul...
- The classifier built from a data set with a highly skewed class distribution generally predicts the more frequently occurring classes much more often than the infrequently occurr...
Following (Blitzer et al., 2006), we present an application of structural correspondence learning to non-projective dependency parsing (McDonald et al., 2005). To induce the corre...