Using Clustering to Improve Classifiers



Previously, I worked on building a classifier in hopes to classify the liquor type involved in a sale between Iowa wholesaler to the retailer based on the data of the sale, volume, and location. You can see the results of that classifier here.

In hopes to improve the performance of the classifier, a team and I looked to build a clustering algorithm to locate new variables that can be used to improve the classifier. See the results of our findings and the performance of our ‘improved’ classifier below.