Chapter 3. Deep Learning for Time Series Modeling
...Yes, it is true that a Turing machine can compute any computable function given enough memory and enough time, but nature had to solve problems in real time. To do this, it made use of the brain’s neural networks that, like the most powerful computers on the planet, have massively parallel processors. Algorithms that run efficiently on them will eventually win out.
Terrence J. Sejnowski (2018)
Deep learning has recently become a buzzword for some good reasons, although recent attempts to improve deep learning practices are not the first of their kind. However, it is quite understandable why deep learning has been appreciated for nearly two decades. Deep learning is an abstract concept, which makes it hard to define in few of words. Unlike a neural network (NN), deep learning has a more complex structure, and hidden layers define the complexity. Therefore, some researchers use the number of hidden layers as a comparison benchmark to distinguish a neural network from deep learning, a useful but not particularly rigorous way to make this distinction. A better definition can clarify the difference.
At a high level, deep learning can be defined:
Deep learning methods are representation-learning1 methods with multiple levels of representation, obtained by composing simple but nonlinear modules that each transform the representation at one level (starting with the raw input) into a representation at a higher, slightly more abstract ...
Get Machine Learning for Financial Risk Management with Python now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.