Publication Details

SELECT * FROM publications WHERE Record_Number=10563
Reference TypeConference Proceedings
Author(s)Calandra, R.; Raiko, T.; Deisenroth, M.P.; Montesino Pouzols, F.
TitleLearning Deep Belief Networks from Non-Stationary Streams
Journal/Conference/Book TitleInternational Conference on Artificial Neural Networks (ICANN)
Keywordsdeep learning, non-stationary data
AbstractDeep learning has proven to be beneficial for complex tasks such as classifying images. However, this approach has been mostly ap- plied to static datasets. The analysis of non-stationary (e.g., concept drift) streams of data involves specific issues connected with the tempo- ral and changing nature of the data. In this paper, we propose a proof- of-concept method, called Adaptive Deep Belief Networks, of how deep learning can be generalized to learn online from changing streams of data. We do so by exploiting the generative properties of the model to incrementally re-train the Deep Belief Network whenever new data are collected. This approach eliminates the need to store past observations and, therefore, requires only constant memory consumption. Hence, our approach can be valuable for life-long learning from non-stationary data streams.
Link to PDF


zum Seitenanfang