A new ERC Grant for research in Artificial Intelligence
Institutional Communication Service
21 July 2017
The Swiss AI Lab at the Dalle Molle Institute for Artificial Intelligence (IDSIA, of USI & SUPSI) confirms its leading role in Informatics research with the new ERC Advanced Grant won by Prof. Jürgen Schmidhuber this past month of June. The project is called AlgoRNN, which stands for Recurrent Neural Networks (RNNs) and Related Machines That Learn Algorithms. Its objective is to go beyond today’s best RNNs, the most general and powerful artificial neural networks. The famous Long Short-Term Memory (LSTM) of Schmidhuber’s team was the first RNN to win machine learning contests, and is now available to billions of users through the world's most valuable public companies, e.g., for speech recognition on over 2 billion smartphones, language translation, etc. Nevertheless, in lots of real-world tasks RNNs do not yet live up to their full potential. Although universal in theory, in practice they fail to learn certain important types of algorithms. This ERC project will address some of the biggest open RNN problems and hottest RNN research topics.
Jürgen Schmidhuber is Full Professor at the USI Faculty of Informatics. His new ERC project will, amongst other things, benefit from the direct access to the fastest supercomputer in the Western world (and the world's 3rd fastest, as of June 2017), with over 5000 GPUs, based in Lugano-Cornaredo, Switzerland.
For further information: http://people.idsia.ch/~juergen/erc2017.html