METODE ANALISIS DISKRIMINAN KUADRAT TERKECIL PARSIAL UNTUK KLASIFIKASI SEGMEN LOYALITAS KONSUMEN SUSU PERTUMBUHAN

Herdina Kuswari (Department of Statistics, IPB University, Indonesia)
Farit Mochamad Afendi (Department of Statistics, IPB University, Indonesia)
Khairil Anwar Notodiputro (Department of Statistics, IPB University, Indonesia)

Indonesian Journal of Statistics and Its Applications
Vol 4 No 2 (2020)

Consumer segmentation is the process of dividing consumers into different segments based on consumer characteristics, making it easier for companies to develop marketing strategies. The segmentation is carried out based on consumer loyalty using the RFM (Recency, Frequency, Monetary) approach a number of 7753 members of a nutritional product loyalty program is considered in the analysis. Partial least square discriminant analysis classification modeling is built using the results of consumer segmentation being the a response variable. The model is not good enough based on the AUC (Area Under Curve) value of the ROC (Relative Operating Characteristic) curve that quite low for each segment. The explanatory variables that have high contribution to the model is X5, X9, and X2 with VIP (Variable Importance in the Projection) values more than 1.