Hands On Data Science Project: Understand Customers with KMeans Clustering in Python

In this walkthrough, we dive into using data science to improve understanding customers by using KMeans clustering to classify online retail customers. Leveraging powerful libraries like sklearn and pandas, we'll walk through the entire process of analyzing customer data to uncover meaningful patterns and segments.

You will learn how to:
- Import and preprocess customer data using pandas.
- Apply KMeans clustering to identify unique customer clusters.
- Visualize the clustering results to gain insights into customer behavior.
- Interpret the outcomes to drive improvement in customer experience.

Whether you're a beginner looking to expand your data science skills or an experienced analyst wanting to refresh your knowledge, this tutorial provides a comprehensive guide to KMeans clustering in Python.

Don't forget to like, subscribe, and hit the notification bell for more data science, data engineering and data analysis content!

#DataScience #KMeans #Clustering #Pandas #Sklearn #Python #CustomerSegmentation #OnlineRetail #DataAnalysis

📖CHAPTERS
00:00:00 - Intro
00:02:35 - Setup
00:05:27 - Exploratory Data Analysis
00:24:23 - Data Cleaning
00:33:20 - How Does KMeans Clustering Work?
00:37:55 - Feature Engineering
01:11:59 - KMeans Clustering
01:25:40 - Cluster Analysis
01:33:31 - Outlier Analysis
01:41:34 - Visualisation
01:46:27 - Outro and Thanks!

🔗LINKS

🖼️PICTURE CREDITS

💵SUPPORT THE CHANNEL

[yohannanpchacko]

Hi! I’ve been working in the industry for the past 23 years, and I really appreciate how your tutorials cover real-world projects, which are often missing in other resources. Great work! I’m eagerly looking forward to your next tutorials, particularly those covering production deployment, operationalization, and automation. Keep up the fantastic work

[kumardipayan]

Hi Bro. I am a beginner in Data Science and have been going through your this video for over 5 days now, for a similar assignment I have to submit. You video was extremely helpful and it was a wonderful experience working on your process on a similar project. It was elaborately, and yet very simply explained. Many Thanks to you

One submission is while the data was not a normal distribution why did we use standardisation (ZScore) for scaling. Ideally we should have used MinMax Scaler and done normalising on skewed distribution. I used MinMax also but didn't get any better results. So I stuck to StdScaler.

Second for the Recency label getting cut off, I searched and found a solution of adjusting the aspect using:-
ax.set_box_aspect(None, zoom=0.9)
Resolved the cutting off problem

Liking and subscribing to your channel, is the minimum I can do to show my gratitude!!

[lucianobv]

what a masterclass trent, i’m speechless!! just wanna thank you for the effort, greetings from Peru😅

[Iri-bp]

I was struggling with data processing while working on my final essay, but the video made everything clear. I'm incredibly grateful for it—thank you so much!

[stevechesney]

Trent, this is one of the best end-to-end data science projects that I've seen, especially for clustering. Thanks for sharing your work and I enjoyed the background music. It helped to keep a good flow and mood going, even as you worked through the tedious parts!

[shadowfox]

EDIT 2: Everyone, just upgrade your modules to the latest version. This should prevent any discrepancies from your plots with the ones you're seeing on the video. This is just a personal note to myself, so I can future proof it lmao.



EDIT 1: Update your Seaborn modules to the latest version. This fixes the problem of the grey violin plot dominating the entire cluster set.

This is the command to update it in a Jupyter notebook: !pip install seaborn --upgrade



Thank you for this hands-on guide, Trent! I did notice however that there were problems with getting the violin plots to appear in the same way as you had in the video. For some reason, when I run it on my computer, the grey violin plot at the very right dominates the entire cluster set. I had to comment out that code line to get the other clusters to show. Even after I downloaded the notebook file you linked in your Github and ran it on my computer it had the same issue. Just thought to point it out.

Thanks anyways! I'll be sure to watch more videos soon.

[harshithasunkara]

Finally completed this video, grateful to come across this project, thanks Trent!

[Libwebs]

I have only watched a minute of this project and I am already excited. Thanks!

[Be_Confident]

I'm from India. That's nice one content I found 😊

[AmoMore-ozuk]

I cannot overstate how much I needed this video, you're calmness comes across well and makes all the information so easily digestible! lol no DNA, just RSA! You definitely have a new subscriber in me!

[markwide]

Hi @TrentDoesMath! - this project was so awesome (!) and there is a lack of these types of step by step walkthrough guids on how to use machine learning libraries in Python on very relatable business problems. Can you make more just like this?? Covering e.g. cost optimization, outlier identification (which can be usefull in all business in one way or the other), and other business problems. I think there is massive interest for these types of walkthrough, and you are very good at it, perfectly balancing the explanation and reasoning without drilling down too deep in the nitty gritty. I think there is a lot of interest here and a testiment is that this video has more views than all of the other combined. Thanks again and hope to see similar ones in the future. Cheers!

[TN-kt]

Very practical and inspiring for a student like me who just learnt some ML methods but found it hard to apply to real world scenario!!Especially for the Exploratory Data Analysis and Feature Engineering are exactly what are missing on the lectures and lack of them makes me out of direction when me really hands on to it. Great appreciation🥰

[christopheanfry]

Amazing video!!! thank you so much for such high quality project and explanation. Just. subscribed. to. your channel. Thanks again

[shahriyarabedinnezhad]

Thank you so much for such a great practice!

[abderrahmaneberriah]

Currently I am practicing each section in your video to get the complete understanding, I love to see more tutos related to daily Data Science Projects, For example how to use Metrics and KPIs to extract insights, Thank you very much for the video.

[lsneerg]

Subbed, awesome work! also much appreciated your sparing of going thru stock codes😂

[Tony-oimw]

Thank you for explaining what's happening at the script-level. Super helpful!!

[-Jannah_Destination]

Very valuable for learners, bro keep making such tutorials

[praveenm]

Its like I have gone thru a year of course curriculum, what an explanation of Data Analysis, i did'nt even noticed the time of the video it just pass by😍👍

[issamsaeed]

Great, thank you. Waiting for the next Hands on project.