Deep Generative Networks For Sequence Prediction

This thesis investigates unsupervised time series representation learning for sequence prediction problems, i.e. generating nice-looking input samples given a previous history, for high dimensional input sequences by decoupling the static input representation from the recurrent sequence representation. We introduce three models based on Generative Stochastic Networks (GSN) for unsupervised sequence learning and prediction. Experimental results for these three models are presented on pixels of sequential handwritten digit (MNIST) data, videos of low-resolution bouncing balls, and motion capture data. The main contribution of this thesis is to provide evidence that GSNs are a viable framework to learn useful representations of complex sequential input data, and to suggest a new framework for deep generative models to learn complex sequences by decoupling static input representations from dynamic time dependency representations.
View on arXiv