Generative Temporal Models with Memory
arXiv: Learning, Volume abs/1702.04649, 2017.
We consider the general problem of modeling temporal data with long-range dependencies, wherein new observations are fully or partially predictable based on temporally-distant, past observations. A sufficiently powerful temporal model should separate predictable elements of the sequence from unpredictable elements, express uncertainty abo...More
Full Text (Upload PDF)
PPT (Upload PPT)