Clustering in R

• Accessed a data set of customer purchasing habits at a winery. Included was demographic data of the customers and their spending habits at this store.
• Cleaned data, removing unnecessary columns and bad entries.
• Scaled dataset based on standardization.
• Selected optimal number of clusters (2) using silhouette scores and elbow plot.



• Utilized K-means clustering to segment, then employed fviz_cluster from the factoextra package to visualize the results.



• Confirmed that the optimal number of clusters had been chosen with a silhouette plot.



• Separated data entries into the two clusters and plotted average and total values for each.



• The two clusters became customer segments for which a marketing plan was made to be recommended to the winery.
• Created charts of purchasing channels and sales of each product category.

Full code can be found in github repository here