Customer Segmentation in Python - PyConSG 2016

preview_player
Показать описание
Speaker: Mao Ting

Description
By segmenting customers into groups with distinct patterns, businesses can target them more effectively with customized marketing and product features. I'll dive into a few machine learning and statistical techniques to extract insights from customer data, and demonstrate how to execute them on real data using Python and open-source libraries.

Abstract
I will go through clustering and decision tree analysis using sciki-learn and two-sample t test using scipy. We will learn the intuition for each technique, the math behind them, and how to implement them and evaluate the results using Python. I will be using open-source data for the demonstration, and show what insights you can extract from actual data using these techniques.

Produced by Engineers.SG

Help us caption & translate this video!

Рекомендации по теме
Комментарии
Автор

Amazing tutorial. Will need to view this a couple of times to get the full understanding of how to combine clustering and supervised learning in one business problem. I have not seen it explained this clearly anywhere. She is one of the best teachers I have seen. Great job and please do more presentations solving a business problem like this. Thanks much!

ramseshadri
Автор

Thank you very much Mao for this great tutorial. You are a very good teacher. I would love to see more videos. But I know this was pycon event. The best thing about your video is the explanation of a business case scenario. This gives a better intuition on how to apply some concept in a real-world problem. Thanks again :)

ThinkGrowIndia
Автор

The 2 sample t test from scipy should be 3 lines of code. Anyone watching this in 2019 should be aware it is much easier now.

TheBjjninja
Автор

Thank you Mao you have made this to look very simple!!

LittleLIAM-uiqy
Автор

At around 33:00, I do not think she fully answered the gentleman’s question regarding assessment of a K means model. You could take all five models to a selected number of domain experts within the company and seek validation on which clustering model made the most sense. In other words the data scientist is not a one man/ woman army. Other managers or experts will need to provide input in this process.

TheBjjninja
Автор

Really great video. Thanks for sharing your knowledge!

snoweepeaks
Автор

I just get a project for customer segmentation by using python :) this video will help me thank you :)

enesteymir
Автор

Is there a possibility to find clear version of video? :(

mediasilv
Автор

may i ask when you define the 3 clusters? could we have a way to let the sklearn to define that?

seymon
Автор

Hi! Off topic question. If I perform the cluster analysis with different algorithms from sklearn how can I check the best one? Thanks in advance!

duiliopaolini
Автор

Nice exposition. Is it possible to get the sample.csv file that you are using?

asanyal
Автор

Great tutorial! can I please have the link to the code as the github link is not working. Thanks!

djibrilkaba
Автор

Wow, she speak clearly, teach easy to understand and be also a beautiful girl. <3

xuanquangnguyen
welcome to shbcf.ru