The Machine Learning (ML) is a technique of artificial intelligence in which a computer acquires the ability to learn without being explicitly programmed and to adapt its program to the changing conditions in which it operates.
Usually, the term is now associated with the modeling of phenomena in predictive function from a large amount of data that describe them and not by an algorithm defined a priori by man.
The systems of machine learning departure receive data related to a particular phenomenon (in a business field, such as the stock market trend in a period of time) and examine them in search of patterns and regularities statistics.
Retroactive feedback to the machine learning
This model is used to test predictions which are then compared with actual data. If the predictions are confirmed, the model is functional and therefore the system is self-programmed to follow it; otherwise, the data is still scanned for others most effective models.
The process is regularly repeated because the algorithm model is verified with new data and can, therefore, adapt to change.
The machine learning applications are the most varied. Achieve a model of a complex system that can perform reliable forecasts and would be the basis for many services ranging from weather forecasts to the analysis of financial markets. It is a sector in which it is working great such as the improvement of medical diagnoses in from the clinical analysis data.
Even without getting to the complete and proper real models, the ability of machine learning to identify regularities and anomalies in large amounts of information is useful to discover fraud in the use of credit cards or the management of online transactions. It is also used in very different fields such as visual recognition or autonomous driving vehicles.
Other examples include machine learning in the recommendation of the automatic system of books or movies according to our preferences and those who suggest new contacts as a function of the network of knowledge we already have.