Data visualization makes correlation analysis easier to use

Correlation analysis It is a basic and important technology in data mining , Is a way to find interesting relationships between variables in large databases . Data analysis Classic case , Wal Mart stores found that customers who buy diapers usually also buy beer , So put beer and diapers together and increase the sales of both . This is a typical application of correlation analysis in the business field , By analyzing a large number of commodity records , Extract useful rules that reflect customer preferences .

With these correlation analyses , Businesses can develop corresponding marketing strategies to increase sales . Relevance analysis is not only widely used in the business field , stay Medical care 、 Education and Finance And other fields have also been effectively applied . Next, I will compile a set of data about the cities that I travel through , Let's learn about the use of correlation analysis .

First, you can find the association analysis component in the text component list , Drag the component onto the canvas .

 

Then copy and paste the table data of the cities where the business trip passes , Click preview to generate . Here's the picture , Figure 1 shows tabular data ; Figure 2 shows the generated association analysis table .

 

  Figure 1

Figure 2

Description of support and confidence

Support : If one of them has low support , It shows that this content appears very few times , And the reference significance of this content is not very big .

Degree of confidence ： The larger the confidence value, the more reliable this item is , The more credible .

Then click properties to view 【 Layer parameters 】, You can see 【 Association analysis configuration 】 Minimum support and minimum confidence in , The minimum support value is 0.3, The minimum confidence is 0.6. The support of Dongguan City in the correlation analysis table is 0.33, From this data we can conclude that , The number of business trips through Dongguan is very small , And the support of Dongguan City is also very low ;

 

Or Dongguan , The example of Fuyun city , contain ｛ Dongguan city, ｝ The contents of the item appear 2 Time , contain ｛ Dongguan City and Fuyun city } The content of item also appears 2 Time , So its confidence level is 1. It shows that the business trip through Dongguan City and Fuyun city is highly reliable .

 

 

Other parameters under layer parameters can be adjusted as required , For example, when there is too much data , You can also manage the display of each page of data , Of course, you can also choose to scroll , Let this big screen be a little more “ flexibility ”.

 

There are also some minor adjustments , These functions are common in life, so I won't introduce them too much , It is worth noting that 【 Progress bar 】, You can change the column numbers into progress bar display , Make the chart more beautiful .

 

Other common properties , The attributes on the right provide more attributes , In order to adjust the ornamental of the components . The attributes in the following figure are also relatively simple 、 frequently-used , I will not introduce too much .

 

Last , Apply correlation analysis to the big data screen , The following figure shows the overall effect of the large screen , It is clear that the associated components are both intelligent and practical , The analysis and mining of data is indispensable .

  The basic use of association analysis component is believed to be well understood , More practical functions still need to be mined and used by yourself , If everyone is right Data visualization If the platform is interested, you can search on the Internet , Have a deep understanding .