Typologies and applications of machine learning, that field of computer science that allows machines to make decisions and predictions
You've read a little about the subject and now you're wondering what machine learning is? Imagine being a professor and one day walking into a classroom of only computing machines. All neat and tidy with apron and pencil case. And then teach them to make decisions autonomously.
What is machine learning? Now to succeed in the intent to be a normal teacher will not help you. To make devices autonomous, in fact, you will have to know the algorithms in the field of machine learning. A related discipline with the, perhaps more talked about, artificial intelligence. Specifically, computational intelligence is a method of data analysis that aims to automatically create analytical models. That is, to allow a computer to process notions, weigh decisions, make them and also predict future options. Obviously these choices must be able to adapt to new situations and not be standard.
Different types of algorithm
(Taken from YouTube)
There are different types of machine learning. Which depend on the use of different algorithms. The three most commonly used are: supervised learning, unsupervised learning, and reinforcement learning. In supervised learning, the system is instructed through the creation of a database with already decoded examples. This allows the machine to catalog what it learns for future use. Based on the experience and patterns it has learned it will then give us predictions. This can be used for prevention in the medical field or to improve the machine's identification of handwriting. In unsupervised learning the system is given a series of examples without any decoding and it will look for commonalities to find a field of similarity. This is the system that search engines use when they point to a series of pages that may be relevant to the given keyword. Reinforcement learning is a machine learning technique that makes the system autonomous by making it learn from its surroundings. This is possible thanks to proximity sensors, GPS and the like. The most concrete example of this is self-driving cars. These learn behavior from the data they receive from sensors: proximity to other cars or pedestrians, speed limits and other road signs. And then they decide accordingly on the action to be taken.
Where is machine learning applied?
As seen from the analysis of algorithms, machine learning can be used in a variety of fields. Uno dei campi dove il machine learnig può essere più funzionale è la pubblicità tracciante. Ovvero quegli annunci pensati in esclusiva per un singolo utente in base al suo profilo e ai suoi comportamenti. Nei social network l’autoapprendimento serve, tra le altre cose, anche per analizzare il sentiment degli utenti in merito a differenti argomenti. E Facebook usa l’apprendimento automatico anche per il riconoscimento facciale nelle foto. Il sistema riconosce il nostro viso apprendendo dalle precedenti immagini dove siete taggati.
Alcuni video presenti in questa sezione sono stati presi da internet, quindi valutati di pubblico dominio. Se i soggetti presenti in questi video o gli autori avessero qualcosa in contrario alla pubblicazione, basterà fare richiesta di rimozione inviando una mail a:[email protected] Provvederemo alla cancellazione del video nel minor tempo possibile.