Deep learning is one of the favourite themes of this blog. It is the most important and powerful tool in machine learning. It’s responsible for many of the biggest breakthroughs in AI, like facial recognition, language translation, or self-driving cars.
The idea behind deep learning is to make it easier to train complex neural networks on large sets of data. Deep neural networks are composed of multiple layers. The first layer takes the input; the last layer produces an output. Each layer consists of several nodes that are connected by weights that represent how much each node should contribute to the output.
On this podcast we talk more about this amazing technology.