Clustering is a way to find common characteristics in a group of data. With clustering we can do partitions in our data, giving to objects similar characteristics to each other and having big dissimilarities with the other clusters. Clustering have such applications in different areas as: marketing, banking, insurance, medicine, etc.
One of the easiest methods is the K-means algorithm. This algorithm divides the data into non-overlapping subsets without any cluster-internal structure or labels. That's how examples within a cluster are very similar, and examples across diffetent clusters are very different.
K-Means try to minimize the intra-cluster distances and maximize the inter-cluster distances.
In this case I have a dataset with different characteristics about some customers.
Here we find:
- Age
- Income
- Level of education (1: high-school, 2: degree, 3: master and 4: PHD)
- Years employed
- Other data
So we want to make groups of people based on some of the characteristics provided by the dataset.
To do that, the code was implementing using libraries such as: pandas, scikit-learn, numpy and matplotlib
This model use the K-Means algorithm with 3 clusters, and provide distributions of customers based on their age, income and education
We can see here three different types of customers:
- Young age and low income (Purple)
- Middle aged and medium income (Yellow)
- Old aged and high income (Blue)
Here is a little bit more difficult to find a real pattern for each customer. That's because the education and age have not necessarily a linear relation to find patterns in between these attributes of the customers
