Neural Networks with Few Multiplications · HackerPapers discussions

Neural Networks with Few Multiplications — mentioned by cs702

Mon, 05 Sep 2016 16:32:17 GMT

As far as I understand this, these guys claim they can train convolutional and many other types of deep neural nets faster by pretraining each layer with a new unsupervised technique via which the layer sort of learns to compress its inputs (a local optimization problem), and then they fine tune the whole network end-to-end with supervised SGD and backpropagation as usual. They have not released code, so no one else has replicated this yet -- as far as I know. If the claim holds, the implication is that layers can quickly learn much of what they need to learn locally, that is, without…

Paper and discussion history. A mention is not an endorsement.

Neural Networks with Few Multiplications — mentioned by secondtimeuse

Mon, 11 Jan 2016 20:14:04 GMT

Deep Neural Networks involve Convolution and Matrix Multiplication operations which have a well defined hard upper bound in terms of speed wrt number of cores, GPUs (~thousands of processors) are in thus indispensable achieving a faster performance. There are newer interesting approaches such as reducing multiplication operations which might improve inference/forward speed even further. http://arxiv.org/abs/1510.03009

Paper and discussion history. A mention is not an endorsement.

Neural Networks with Few Multiplications — mentioned by dharma1

Mon, 09 Nov 2015 23:11:38 GMT

did you see this? http://arxiv.org/pdf/1510.03009v1.pdf

Paper and discussion history. A mention is not an endorsement.

Neural Networks with Few Multiplications — HN submission

Mon, 09 Nov 2015 04:09:38 GMT

Neural Networks with Few Multiplications

66 points; 8 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Neural Networks with Few Multiplications — HN submission

Fri, 16 Oct 2015 23:35:20 GMT

Neural Networks with Few Multiplications (Bengio and Co.)

3 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.