Build segmentation with k-means clustering

In this exercise, you will build the customer segmentation with KMeans algorithm. As you've identified in the previous step, the mathematically optimal number of clusters is somewhere around 3 and 4. Here, you will build one with 4 segments.

The pre-processed dataset has been loaded as wholesale_scaled_df. You will use it to run the KMeans algorithm, and the raw un-processed dataset as wholesale - you will later use it to explore the average column values for the 4 segments you'll build.

This exercise is part of the course

Machine Learning for Marketing in Python

View Course

Exercise instructions

Import the KMeans algorithm from sklearn.cluster module.
Initialize KMeans algorithm with 4 clusters and a random state set to 123.
Fit the model on the pre-processed wholesale_scaled_df dataset.
Assign the generated labels to a new column called segment in the raw wholesale dataset

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

# Import `KMeans` module
from sklearn.cluster import ___

# Initialize `KMeans` with 4 clusters
kmeans=KMeans(___=4, random_state=123)

# Fit the model on the pre-processed dataset
kmeans.fit(___)

# Assign the generated labels to a new column
wholesale_kmeans4 = wholesale.assign(segment = kmeans.___)

Edit and Run Code