RMSS 2017

​

After an introduction to Big Data and Neural Networks, the presentation  will be focused on two particular type of neural network: one supervised (convolutional neural network) and the other unsupervised (BCM neural network). The Convolutional Neural Network will be applied to a specific problem, called super-resolution analysis, that permit to obtain high resolution signals (including spectra and images) from low resolution samples. The BCM neural network will be illustrated as a unsupervised method aimed to extract features from complex signals. Finally, it will be discussed a new finding about BCM in the case of linear dependent input.