https://github.com/dmarks84/ind_project_mall-customer-clustering--kaggle

Independent Project - Kaggle Dataset-- I worked with the Mall Customer Segmentation Dataset, which provided a various instances of shoppers of different ages, incomes, etc. I utilized unsupervised ML clustering algorithms to identify useful customer segments.
https://github.com/dmarks84/ind_project_mall-customer-clustering--kaggle

clustering dataframes dbscan kmeans-clustering market-segmentation mean-shift pandas python sklearn technical-analysis technical-communication unsupervised-ml

Last synced: 8 days ago
JSON representation

Independent Project - Kaggle Dataset-- I worked with the Mall Customer Segmentation Dataset, which provided a various instances of shoppers of different ages, incomes, etc. I utilized unsupervised ML clustering algorithms to identify useful customer segments.

• Host: GitHub
• URL: https://github.com/dmarks84/ind_project_mall-customer-clustering--kaggle
• Owner: dmarks84
• License: bsd-3-clause
• Created: 2024-02-12T19:58:53.000Z (11 months ago)
• Default Branch: main
• Last Pushed: 2024-02-27T22:39:11.000Z (10 months ago)
• Last Synced: 2024-11-05T19:22:24.691Z (about 2 months ago)
• Topics: clustering, dataframes, dbscan, kmeans-clustering, market-segmentation, mean-shift, pandas, python, sklearn, technical-analysis, technical-communication, unsupervised-ml
• Language: Jupyter Notebook
• Homepage: https://www.kaggle.com/code/danpmarks/mall-customer-clustering
• Size: 1.36 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Ind_Project_Mall-Customer-Clustering--Kaggle

## Screenshot

![screenshot](https://github.com/dmarks84/Ind_Project_Mall-Customer-Clustering--Kaggle/blob/main/mall_screenshot.png)

## Summary

I worked with the Mall Customer Segmentation Dataset, which provided a various instances of shoppers of different ages, incomes, etc.  I first reviewed the data and used paired plots to determine that gender was not a greaty differentiator/indicator of customer segmentation.  I then employed three different Scikit-learn unsupervised clustering algorithms: K-Means, DBSCAN, and Mean Shift to determine if customer segmentation as a function of the other features-- Age and Income-- was possible and/or clear.  With K-Means, I plotted inertia for different numbers of clusters to hone in on potential elbow points, and I ran through several combinations of min_samples and epsilon for DBSCAN.

## Results

Both K-Means and Mean Shift settled on a cluster number of 5 when looking at the Score as a function of Income.  They overlapped substantially in their assignment of segment/cluster, leading to clear indication behind the "meaning" of each segment:

 - **Label 0** is "mid" income and "mid" spending

 - **Label 1** is "high" income and "low" spending

 - **Label 2** is "high" income and "high" spending

 - **Label 3** is "low" income and "low" spending

 - **Label 4** is "low" income and "high" spending

## Skills (Developed & Applied)

Programming, Python, Pandas, Dataframes, Unsupervised-ML, Clustering, DBSCAN, K-Means, Mean Shift, Scikit-learn, Technical Communication