The clustering has provided data analysis in many contexts of Computer Science. It is widely applied in Ambient Intelligence and Ubiquitous Computing for information processing, with geolocation data prominently. In this paper, we introduce a dynamic fuzzy temporal clustering algorithm (DFTC) to detect stays of users in urban environments based on locations from imprecise sensors. Our approach includes fuzzy evaluation of temporal and probabilistic data providing analysis in real time. As results, we have developed a mobile application which integrates the DFTC and detects satisfactorily user stays related to urban commerces from a real environment.

Zero-shot classification (ZSC) is the task of learning predictors for classes not seen during training. Although the different methods in the literature are evaluated using the same class splits, little is known about their stability under different class partitions. In this work we show experimentally that ZSC performance exhibits strong variability under changing training setups. We propose the use ensemble learning as an attempt to mitigate this phenomena.

No abstract

No abstract