How Does Deep Learning Work?
A new theory seeks to explain the success of Deep Learning by highlighting generalization as a kind of algorithmic ‘forgetting.’
As an example, some photos of dogs might have houses in the background, while others don’t. As a network cycles through these training photos, it might “forget” the correlation between houses and dogs in some photos as other photos counteract it. It’s this forgetting of specifics, Tishby and Shwartz-Ziv argue, that enables the system to form general concepts. Indeed, their experiments revealed that deep neural networks ramp up their generalization performance during the compression phase, becoming better at labeling test data.