Neural networks are a powerful and highly flexible class of learning method, able to learn highly complex and non-linear functions. As their name suggests, neural networks are loosely inspired by the structure of the brain, mimicking the way biological neurons signal to one another. They are typically organized into layers, each transforming the activation from the previous layer and passing it on to the next. These transformations amount to a composition of simple, but non-linear functions. Through these transformations they are able to represent the data in such a way that it is more amenable for the task. When many such layers are stacked together, the learned representations can become very powerful, and we call this deep learning.
4.1 Introduction to Neural Networks
4.4 Architectures and Innovations
<aside> ⬅️ ‣
</aside>
<aside> ➡️ 4.1 Introduction to Neural Networks
</aside>