Deep Learning is a discipline of artificial intelligence, and more precisely a branch of Machine Learning. It is about letting machines learn from their experiences, just like humans do.
What’s the difference between machine learning and deep learning?
The difference with Machine Learning is that Deep Learning algorithms don’t really have a limit in terms of learning ability. The more data they receive to train on, the more these systems improve their performance.
The artificial neural networks on which deep learning is based learn by discovering structures in the data provided to them. These networks develop computational models composed of multiple processing layers, in order to create multiple levels of abstraction to represent the data.
A convolutional neural network type, deep learning model can be trained on millions of images of cats, for example. It will then learn to recognize the pixels representing a cat, in order to classify them as such.
With a conventional machine learning system, a human expert would have to spend a considerable amount of time setting up the system to enable it to detect the characteristics of the cat.
In the case of Deep Learning, it suffices to provide it with a large quantity of images.
The performance of deep learning algorithms surpass that of machine learning systems in a large number of tasks including computer vision, speech recognition and robotics. However, it is necessary to configure the “hyperparameters” for the model to be effective.
What is deep learning used for?
Deep Learning is used for a wide variety of applications such as disease detection, inspection of industrial equipment, autonomous vehicles, discovery of exoplanets or drugs, study of the genome or even the fight against global warming.