Learn More
Distributed computing is essential for handling very large datasets. Online learning is also promising for learning from rapid data streams. However, it is still an unresolved problem how to combine them for scalable learning and prediction on big data streams. We propose a general computational framework called loose model sharing for online and(More)
Aging is associated with changes in physical characteristics and decline of many physiological functions. The aging process have been described by various theories, in particular the free radical theory of aging has received widespread attention. It has been accepted that the oxidative stress or damage induced by free radicals is related to aging. In this(More)
Low-variance gradient estimation is crucial for learning directed graphical models parameterized by neural networks, where the reparameterization trick is widely used for those with continuous variables. While this technique gives low-variance gradient estimates, it has not been directly applicable to discrete variables, the sampling of which inherently(More)
Learning discrete representations of data is a central machine learning task because of the compactness of the representations and ease of interpretation. The task includes clustering and hash learning as special cases. Deep neural networks are promising to be used because they can model the non-linearity of data and scale to large datasets. However, their(More)
  • 1