Multiple data sources containing different types of features may be available for a given task. For instance, users’ profiles can be used to build recommendation systems. In a...
Learning by imitation has shown to be a powerful paradigm for automated learning in autonomous robots. This paper presents a general framework of learning by imitation for stochas...
Abstract. The goal of predictive toxicology is the automatic construction of carcinogenecity models. Most common artificial intelligence techniques used to construct these models ...
In this paper we propose a random set framework for learning linguistic models for prediction problems. We show how we can model prediction problems based on learning linguistic p...
Abstract. We investigate the generalization behavior of sequential prediction (online) algorithms, when data are generated from a probability distribution. Using some newly develop...