welcome to our second issue. This week we are regaling you with a veritable cornucopia of interesting developments in deep learning from emoji embeddings over facebooks newly open sourced fastText library to what to do when your gradient descent doesn’t converge.
If you like this issue follow us on twitter, tweet about this issue, tell your colleagues, your spouse, the people in front of you in line at the grocery store… really we are grateful for any word that gets out.
Dango is a floating assistant that runs on your phone and predicts emoji, stickers and GIFs based on what you and your friends are writing in any app. It does so using an RNN which creates word (emoji) embeddings.
A detailed overview of how NVIDIA used convolutional neural networks (CNNs) to map the raw pixels from a front-facing camera to the steering commands for a self-driving car. A more rigorous presentation can be found in the authors’ eponymous paper.
The paper briefy reviews several recent results on hierarchical architectures for learning from examples, that may formally explain the conditions under which Deep Convolutional Neural Networks perform much better in function approximation problems than shallow, one-hidden layer architectures.
Deep Learning Weekly aims at being the premier news aggregator for all things deep learning. We keep tabs on major developments in industry be they new technologies, companies, product offerings or acquisitions so you don't have to. In addition our 'Learning' section features new content that makes difficult to understand areas in deep learning accessible to a wider audience and our 'Papers & Publications' section brings you the most exicting new research. If you have any thoughts or ideas how we might improve this newsletter we are interested in hearing them. Thank you in advance!
We use third party cookies and scripts to improve the functionality of this website.Accept