ARTIFICIAL INTELLIGENCE

Artificial intelligence consists in implementing a certain number of techniques aimed at allowing machines to imitate a real (human) form of intelligence.

From Google to Microsoft, Apple, IBM or Facebook, all the major companies in the IT world are working today on the problems of artificial intelligence by trying to apply it to a few specific fields. Each of them has set up artificial neural networks made up of servers and allowing to process heavy calculations within gigantic databases.

MACHINE LEARNING

Machine Learning is a system in which input data is chosen and implemented in an algorithm in order to have a desired prediction in output.

• Giving computers the ability to learn without being explicitly programmed.
• Designing algorithms that can learn from patterns in the data and exploit them
• A subset of artificial intelligence where the machine is trained to learn from past experience. Past experience is developed through the data collected. The more experience the model has (trained itself), the more effective it will be at making predictions.

BIG DATA PROCESSING

Data processing is undoubtedly the most important part when talking about artificial intelligence. Indeed, a model as complex as it is, is useless without a large amount of data to train it. Data are very complicated to handle, indeed, they must be in large quantity but also relevant and this requires a lot of processing work.

NEURAL NETWORKS

Deep Learning is a type of artificial intelligence derived from Machine Learning where the machine is able to learn by itself, unlike programming where it simply executes predetermined rules to the letter.

Deep Learning is based on a network of artificial neurons inspired by the human brain. This network is composed of dozens or even hundreds of “layers” of neurons, each of which receives and interprets information from the previous layer.

At each stage, the “bad” answers are eliminated and sent back to the upstream levels to adjust the mathematical model. As it proceeds, the program reorganizes the information into more complex blocks. When this model is later applied to other cases, it is normally able to recognize the desired concept without anyone having ever told it or taught it. The starting data is essential: the more different experiences the system accumulates, the better it will perform.