One of the main drawbacks of the Semantic Web is the lack of semantically rich data, since most of the information is still stored in relational databases. In this paper, we presen...
The bounding box representation employed by many popular object detection models [3, 6] implicitly assumes all pixels inside the box belong to the object. This assumption makes th...
Abstract: We investigate the structure of model selection problems via the bias/variance decomposition. In particular, we characterize the essential structure of a model selection ...
Design of collaborative learning (CL) scenarios is a complex task, but necessary if the goal of the collaboration is learning. Creating well-thought-out CL scenarios requires exper...
We evaluate two broad classes of cognitive mechanisms that might support the learning of sequential patterns. According to the first, learning is based on the gradual accumulation...