Mathematics Colloquia and Seminars
From machine learning to brains and back againOptimization
|Speaker:||David Rolnick, MIT|
|Location:||3106 Math Science Building|
|Start time:||Tue, Oct 18 2016, 4:10PM|
Recent work in deep learning, combined with cutting-edge microscope technology, has made it possible to reconstruct networks of neurons in the brain with unprecedented accuracy and scale. I will discuss algorithms involved in mapping the brain and why traditional deep learning may not be enough. I will also show how understanding the "connectome" (the brain's neural network) can lead to new insights in artificial neural networks.
Contact Jesus De Loera email@example.com if you wish to meet the speaker who will be here already Monday.