Unreasonable Effectiveness
Understand the surprising power of deep neural networks to recognize complex patterns and perform human-like tasks. Learn how layers in deep learning models create abstractions from basic shapes to high-level concepts, eliminating the need for manual feature engineering and enabling advancements in AI applications like medical diagnosis and facial recognition.
Unreasonable effectiveness of deep networks
Throughout this course, we might have been surprised by the capabilities of simple programs like our first MNIST classifier. And yet, little prepares us for the uncanny capabilities of modern deep networks. In the words of one famous researcher, those networks are “unreasonably effective.” In a sense, they do nothing more than recognize patterns yet, they increasingly beat us at quintessentially human tasks, like facial recognition or medical diagnosis.
Note: The short answer is that we don’t understand why deep neural networks work so well on so many tasks. Indeed, a lot of ongoing research is bent to explain that fact.
At first sight, it’s not even obvious why deeper networks work better than more shallow ones. A famous theorem from 1989, called the universal approximation theorem, proves that with enough hidden ...