When the number of labeled examples is limited, traditional supervised feature selection techniques often fail due to sample selection bias or unrepresentative sample problem. To ...
Computer manufacturers spend a huge amount of time, resources, and money in designing new systems and newer configurations, and their ability to reduce costs, charge competitive p...
Abstract— Automatic pattern classifiers that allow for incremental learning can adapt internal class models efficiently in response to new information, without having to retrai...
— This paper proposes a method for identifying an object which contains an accelerometer out of many moving objects in the view of a stationary camera using motion data obtained ...
Gene network reconstruction is a multidisciplinary research area involving data mining, machine learning, statistics, ontologies and others. Reconstructed gene network allows us t...