=================================== TCN: Temporal Convolutional Network

This is a generic Temporal Convolutional Network implementation as proposed in An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling <https://arxiv.org/pdf/1803.01271.pdf>_ by Bai et al. developed with Tensorflow.

The implementation has been tested on the following task contained in this repository:

• Sequential MNIST <https://github.com/jakeret/tcn/blob/master/mnist.py>_
• Copy memory <https://github.com/jakeret/tcn/blob/master/copy_memory.py>_
• Text classification on the IMDB movie dataset <https://github.com/jakeret/tcn/blob/master/imdb.py>_

Architectural elements in a TCN. (a) A dilated causal convolution with dilation factors d = 1, 2, 4 and filter size k = 3. (b) TCN residual block. (c) An example of residual connection in a TCN

.. image:: https://raw.githubusercontent.com/jakeret/tcn/master/docs/tcn_architecture.png :alt: Temporal Convolutional Network :align: left

The corresponding elements in TensorBoard

.. image:: https://raw.githubusercontent.com/jakeret/tcn/master/docs/tcn.png :alt: Temporal Convolutional Network :align: left