Data Mining and machine learning

Machine Learning:

It is the core research area of artificial intelligence. Currently, it is defined:Use experience to improve computer system performance. For "experience", in fact, "experience" exists in the form of data in the computer, so Machine LearningData needs to be analyzed and used.

 

Improving generalization ability is one of the most important issues in machine learning. The generalization ability represents the ability of the machine learning system to adapt to new events. Simply put, the higher the generalization ability, the more accurate the system makes predictions on events.

 

Data mining:

"Data Mining" and "Knowledge Discovery" are generally considered the same. In many cases, there are alternative terms.

Data mining as its name implies: Find useful knowledge from massive data. Data mining can be considered as a cross-application of machine learning and databases. It uses machine learning technology to analyze massive data and database technology to manage massive data.

There is also "Statistics". Many algorithms of Statistics usually need further research through machine learning to apply them to data mining.

From the perspective of data analysis, most data mining technologies are applied to machine learning, but we cannot think that data mining is an application of machine learning. Traditional machine learning does not take massive data as research and processing objects. Many technologies

It is only applicable to small and medium-sized data. If these technologies are applied to massive data, the results will be very bad. Therefore, data mining also requires specialized transformation of these technologies.

For example, "Decision Tree" is a good machine learning technique that not only has strong generalization ability, but also has understandable learning results. The traditional approach is to read all the data into the memory for analysis. However, it is obviously not feasible for massive data volumes. This requires processing,

For example, by introducing efficient data structures and scheduling policies.

In addition, as an independent discipline, data mining has its own unique things. For example,Association Analysis". In short, association analysis is to find out a strange but meaningful Association like beer and diapers from a lot of data. If 20 out of 100 purchased diapers and 16 out of the 20 purchased diapers bought beer, you can write it as "diapers → beer [Support = 20%, confidence Level = 80%.