Month: February 2017

 ?Deep Learning is presented as Energy-Based Learning Indeed, we train a neural network by running BackProp, thereby minimizing the model error–which is like minimizing an Energy. But