PCA for CRM data

Machine Learning for Marketing Analytics in R

Verena Pflieger

Data Scientist at INWT Statistics

Machine Learning for Marketing Analytics in R

PCA helps to...

  • Handle multicollinearity
  • Create indices
  • Visualize and understand high-dimensional data
Machine Learning for Marketing Analytics in R

Data for PCA

str(dataCustomers, give.attr = FALSE)

Classes 'tbl_df', 'tbl' and 'data.frame':    989 obs. of  16 variables:
 $ nOrders            : int  104 17 5 18 21 2 18 12 14 7 ...
 $ nItemsOrdered      : int  138 21 6 27 41 2 29 14 19 13 ...
 $ nItemsSold         : int  66 4 3 3 35 1 11 11 9 2 ...
 $ salesOrdered       : num  37813 10653 1226 31529 17935 ...
 $ salesSold          : num  18031 1500 759 3803 14246 ...
 $ returnRatio        : num  0.522 0.81 0.5 0.889 0.146 ...
 $ shareOwnBrand      : num  0.54 0.48 1 0.15 0.63 0 1 1 0.42 0.31 ...
 $ shareSale          : num  0.52 0.67 0.17 0.19 0.12 0 0.28 0.07 0.37 ...
 $ shareVoucher       : num  0.09 0.1 0.5 0.07 0 0 0.52 0.29 0.16 0 ...
 $ crDuration         : int  1472 1506 1453 1340 1449 749 997 1513 1499 ...
 $ monetaryReturnRatio: num  0.523 0.859 0.381 0.879 0.206 ...
 $ meanDaysBetwOrders : int  14 94 363 79 72 749 59 138 115 254 ...
 ...
Machine Learning for Marketing Analytics in R

Correlation structure

library(corrplot)
dataCustomers %>% cor() %>% corrplot()

Machine Learning for Marketing Analytics in R

Let's practice!

Machine Learning for Marketing Analytics in R

Preparing Video For Download...