Let's Cluster some Credit Cards with BIRCH!

Here, I use birch clustering algorithm.

Explain:

In this notebook I'll try to cluster some credit cards. we go through preprocessing, choose a model and train it on our data, and then evaluate and visualize our outcomes.

Dataset

I used Credit Cards dataset for this project. You can download the dataset here

A bout dataset features

• CUSTID : Identification of Credit Card holder (Categorical)
• BALANCE : Balance amount left in their account to make purchases
• BALANCEFREQUENCY : How frequently the Balance is updated, score between 0 and 1 (1 = frequently updated, 0 = not frequently updated)
• PURCHASES : Amount of purchases made from account
• ONEOFFPURCHASES : Maximum purchase amount done in one-go
• INSTALLMENTSPURCHASES : Amount of purchase done in installment
• CASHADVANCE : Cash in advance given by the user
• PURCHASESFREQUENCY : How frequently the Purchases are being made, score between 0 and 1 (1 = frequently purchased, 0 = not frequently purchased)
• ONEOFFPURCHASESFREQUENCY : How frequently Purchases are happening in one-go (1 = frequently purchased, 0 = not frequently purchased)
• PURCHASESINSTALLMENTSFREQUENCY : How frequently purchases in installments are being done (1 = frequently done, 0 = not frequently done)
• CASHADVANCEFREQUENCY : How frequently the cash in advance being paid
• CASHADVANCETRX : Number of Transactions made with "Cash in Advanced"
• PURCHASESTRX : Numbe of purchase transactions made
• CREDITLIMIT : Limit of Credit Card for user
• PAYMENTS : Amount of Payment done by user
• MINIMUM_PAYMENTS : Minimum amount of payments made by user
• PRCFULLPAYMENT : Percent of full payment paid by user
• TENURE : Tenure of credit card service for user

Essential libraries that I used :

• numpy
• pandas
• matplotlib
• seaborn
• Scikit-learn

Preprocessing

1. remove the outliers
2. impute missing data
3. scale the data
4. Reduce dimentions using PCA

check for outliers

Using IQR, we can follow the below approach to find outliers:

• Calculate the first and third quartile (Q1 and Q3).
• Further, evaluate the interquartile range, IQR = Q3-Q1.
• Estimate the lower bound, the lower bound = Q1*1.5
• Estimate the upper bound, upper bound = Q3*1.5
• The data points that lie outside of the lower and the upper bound are outliers.

Removing the outliers

first, let's get rid of the noise. we're going to first set all outliers as NaN, so it will be taken care of in the next stage, where we impute the missing values.

Imputing the missing data

I use KNN imputer: Each sample’s missing values are imputed using the mean value from n_neighbors nearest neighbors found in the training set.

Scale the Data

I use StandardScaler

Dimention Reduction using PCA

K-means, DBSCAN and agglomerative clustering, all use the Euclidean distance, which starts to lose its meaning when the number of dimensions starts increasing. so, before using these methods, we have to reduce the number of dimensions. I'm going to use PCA, which is by far the most popular dimensionality reduction algorithm.

If you are not familiar with PCA or need to learn more about it, I highly recommend you read Here on this dimentionality reduction method. here I set parameter n_components equals to 2 for better visualizing the results.

Optimize number of clusters

I used Elbow method for estimating the optimize number of clusters. If you want to know more about it click Here.

And I also used Silhouette method for this!

Birch clustering

At the end, I used Birch(Balanced Iterative Reducing and Clustering using Hierarchies) for clustering. If you want know more about it click Here