Using the k-means algorithm for zoning real estate pricing zones

This publication does not belong to the materials of the series "here it is event horizon", but on the contrary, as an advisor on the use of recognized methods of BigData analysis in the practical activities of ordinary people far from the zoo with Python (Python), Escuels (SQL), Siplusplus ( C ++) and others - appraisers when determining the market value of real estate. The need to determine the impact of location on the value of real estate is beyond doubt. This fact is enshrined in practice, in the requirements of FSO-7 (Federal standard of valuation "Real estate appraisal (FSO N 7)" paragraphs 11b and 22e.





 At the moment, there are such “heat maps of real estate” in the network, but they are narrow in purpose, as they reflect the cost of apartments, and other types of real estate need to be assessed as well. On the other hand, these information sources are not endowed with the necessary completeness, which limits their use in forensic examination.  





Of course, each appraiser knows his territory and has his own "heat map". I can imagine how I myself carried out regionalization without the use of mathematical methods (Fig. 1). 





. 1. 2019 . ( « № 01/-2019 ( ) - — » . _\2. \2.1 ». 19.11.2019 № 19- « ( ) - – »





  g-means 2021, . 2.





. 2. 2021 . g-means ( )





 - « » etagi.com (.3) .





. 3. etagi.com ( 2021 .) ( )





 





, « » , 80% , . , , .





  « » - (. 4) .





. 4. . . ( 2021 .). ( ).





  «-» - . . 5 .





. 5. . «-»





  . , . - , - . , .





  «-» .





  «Etagi.com» .





  . , 6 7 56214-2014 « . 1. » .67. « » , .88. « » , .71. « » , , .





  : ) ( ); ) ; ) ( ).





  , , .





  , 99% , . «_1 v.s. _2». , .





  , k-means (g-means) , , Excel’ , , , . . «Loginom Company» Deductor Academic .





1. ( ). « » Excel, .





2. .





3. .





  , Deductor Academic.





  : .





  : .





  , , .





  . 2 . Deductor Academic , . .2 ( ), . , 6.





. 6. ( / ).





  , . 7.





. 7. .





  . , 10%. «_1 v.s. _2». 7 969,14, 7 873,28. . Deductor Academic .





  ? . , . 237- k-means (g-means). . .





 





.





 





«», , .





  , . . (.11 , -7), 40-100 , , .





  - , . ( ) - 60.957875, 76.878470. Excel .8.





. 8.





  . . . « XLSX». .





  . , , , . , – , , , , , – .





  . Deductor Academic ,





1) ;





2) , ;





3) ( Excel' , ).





9.





. 9. ) – , . ) . ) 8, 9 - .





  . 10, :





) «1»;





) «_» ;





) « ( ) (*.txt);





) ( Deductor Academic , , . , ( / ) , ).





. 10.





  Deductor Academic.





. 11.





  12 , , Deductor Academic .





. 12. Deductor Academic ()





  13 Deductor Academic , .





. 13. Deductor Academic.





  14 .





. 14. Deductor Academic.





 





.





 





.15 .





.15. Deductor Academic.





  (Data minig) … , .16. , . Deductor Academic , .





. 16.





  .





. 17. .





  17 , (0,5512), (0,6898). (.18).





. 18. . ( - , – )





  . , .





. 19. .





. 20. .





Fig. 21. Set the number of clusters with g-means, the number of clusters is determined automatically.





Fig. 22. Process start menu.





Fig. 23. List of output parameters.





Fig. 24. Analysis results.





 





The results on Yandex.Map are presented here and in Figure 25.





Fig. 25. Comparison of the results of the analysis in Deductor Academic with the analysis that was carried out when performing this task by other mathematical means.





 





And at the end - Welcome to the embrace of mining date.








All Articles