Project 4: Mall Customer Research
This research is based on mall goers and their different traits and how well they shop at the mall. Specifically, we try to answer 2 questions with this research; Do mall goers with a higher income tend to have a higher spending score than a mall goer with a lower income? Does gender equate to a higher spending score, do women tend to have a higher spending score than men or vice versa? To answer these questions, we will use the clustering method for data analysis.
What is Clustering?
Clustering is the sorting of data points into several groups. Determining which groups data is sorted into depends on the similarties of each data point.
The Data
The dataset being used for this research is from Kaggle, and is titled Mall Customer Segmentation Data created by Vijay Choudhary.
Data Understanding/Visualization
Looking at this heat map chart we can see a relevantly fair correlation between Annual Income and Spending Score.
Analyzing the plot we can see that the data is very spread out and it shows that most people that make more spend more and those that make less spend less, the plot also showed that gender really is not a factor as men and women like to shop. We will look into this more with clustering.
Pre - Processing
Our data set will not need to be pre-processed as it is fine as it is at this current time. One reason being that the dataset is fairly small.
Modeling
We chose K-Means clustering to look at how Annual Income and Spending Scores relate, looking at the plot we can see there are 5 distinct groups, these represent the different earning groups compared to how high their relative spending score is.
Looking at the plot we can see that gender does not equate to a higher spending score as men and women are all over the place when it comes to spending.
Storytelling
The modeling shows us that annual income does not impact your spending when visiting the mall as the average spending score is 50. However, most people tend to shop at a level they can afford. It also showed that gender does not impact a spending score as both men and women have mixed spending scores when going to the mall.
Impact
This research can show us that malls are still important for every consumer as online shopping continues to boom. This data can show us what kind of people still go to the mall and the mall stores can tailor their stores to the people that use the mall. But it can also mean bad news for some stores as the people who shop at their stores might turn to online shopping.
References
-
Image from Google Images
-
https://blog.floydhub.com/introduction-to-k-means-clustering-in-python-with-scikit-learn/
-
in-class Pokemon Clustering Activity