Relational instance-based learning (RIBL) algorithms offer high prediction capabilities. However, they do not scale up well, specially in domains where there is a time bound for classification. Nearest prototype approaches can alleviate this problem, by summarizing the data set in a reduced set of prototypes. In this paper we present an algorithm to build Relational Nearest Prototype Classifier...

The main objective of this deliverable is to provide a set of proof-of-concept prototypes to illustrate some particular solutions developed within CORTEX. Additionally , these proof-of-concept prototypes should also be seen as technology demonstrators , which will provide relevant input for the construction of integrated CORTEX demo applications (see deliverable WP4-D8). More specifically, this...

The object-oriented programming language design space consists of class-based and prototype-based languages. Both language families have been shown to posses many advantages but also several disadvantages with respect to software construction. Hybrid languages featuring both prototype-based and class-based mechanisms have been proposed as a solution. Unfortunately these languages not only unify...

OBJECTIVES There is still limited understanding of how different kinds of drinker prototypes are associated with adolescent drinking. This study uses the strengths of multiple time-point diary measures (enhanced validity of alcohol use measurement) to test the predictive value of abstainer, moderate and heavy drinker prototypes in social situations. We examined whether the favorability of these...