What is data mining?

 1、 The concept of data mining

   Data mining is from a large number of 、 Not completely 、 Noisy 、 Vague 、 In random actual data , Extract the implied , What people don't know in advance , But the process of potentially useful information and knowledge .

  The data source used for data mining must be real and massive , And may be incomplete and include some interference data items . The discovered information and knowledge must be of interest and usefulness to the user . In general , The result of data mining is not required to be completely accurate knowledge , It's about finding a big trend .

  Data mining can be simply understood as the operation of a large amount of data , The process of discovering useful knowledge . It is an interdisciplinary subject covering a wide range of fields , Including machine learning 、 mathematical statistics 、 neural network 、 database 、 pattern recognition 、 Rough set 、 Fuzzy mathematics and other related technologies .

  2、 Application of data mining

  For specific applications , Data mining is a process of using various analysis tools to find the relationship between models and data in massive data , These models and relationships can be used to make predictions .

  Knowledge discovery in data mining , It is not to discover the truth that is universal , Nor is it to discover new natural science theorems and pure mathematical formulas , It is not a machine theorem proving . actually , All found knowledge is relative , There are specific premises and constraints , Oriented to a particular field , At the same time, it should be easy for users to understand , It is better to express the findings in natural language .

  Data mining is actually a kind of deep-seated data analysis method . Data analysis itself has a long history , But in   In the past , The purpose of data collection and analysis is for scientific research . in addition , Due to the limitation of computing power at that time , Complex data analysis methods that analyze large amounts of data are greatly limited .

  3、 Value types of data mining

   Data mining is to find valuable data in the mass of data , Provide basis for business decision-making . Value usually includes relevance 、 Trends and characteristics .

 1） The correlation

  Correlation analysis refers to the analysis of two or more variable elements with correlation , So as to measure the   Degree of correlation .

  Correlation analysis can only be carried out when there is a certain relationship or probability between elements . Correlation is not causality , The scope and field covered almost every aspect we have seen . Correlation analysis is used to determine changes between data , That is, whether the change of one or several attributes will affect other attributes , What is the impact . chart 1 These are examples of several common correlations .

2） trend

Trend analysis refers to the results that will actually be achieved , Compare with the historical data of similar indicators in the financial statements of different periods , To determine the financial position 、 An analytical method for the change trend and law of operating results and cash flow . The trend and trend of data can be predicted through the line chart , It can also be achieved through the link comparison 、 The results of the comparison are explained in a year-on-year manner , Pictured 2 Shown .

3） features

Feature analysis refers to finding the features of the main objects according to the specific analysis contents . for example , Internet data mining is to find out all aspects of the characteristics of users to portrait users , And according to different users, the user group will be labeled accordingly . Pictured 3 Shown .