In real world scenes, objects to be classified are usually not visible from every direction, since they are almost always positioned on some kind of opaque plane. When moving a cam...
Abstract. Successful multi-target tracking requires locating the targets and labeling their identities. This mission becomes significantly more challenging when many targets freque...
We investigate the performance of two machine learning algorithms in the context of antispam filtering. The increasing volume of unsolicited bulk e-mail (spam) has generated a nee...
Ion Androutsopoulos, Georgios Paliouras, Vangelis ...
In this investigation, we propose a probabilistic approach for estimating the ages of Blog authors by means of Naive Bayesian Classifier. We can learn context of characteristic wor...
The basic nearest neighbour classifier suffers from the indiscriminate storage of all presented training instances. With a large database of instances classification response time ...