OVERVIEW: This project involves analyzing customer data to perform segmentation based on income and spending patterns. The insights are visualized in a Power BI dashboard, and clustering is implemented using an ML model in Google Colab.
TOOLS AND TECHNOLOGIES-
- Google Colab: Building and training the machine learning model
- Python: Used for data cleaning, processing and clustering
- Power BI: Interactive data visualization
PROJECT WORKFLOW- This project involves multiple stages, from data cleaning to machine learning and visualization. Below is the complete workflow:
- Data Cleaning and Preprocessing:
- The original dataset was cleaned and processed to handle missing values, remove duplicates, and ensure data consistency.
- Features like Annual Income, Spending Score, and Gender were prepared for analysis.
- Clustering with K-Means:
- K-Means clustering was applied in Google Colab to segment customers into distinct groups based on income and spending patterns.
- The optimal number of clusters was determined using the Elbow Method.
- Clustered Data Export:
- The processed dataset, enriched with cluster labels, was exported as a CSV file for further visualization.
- Dashboard Creation in Power BI:
-
The processed dataset was imported into Power BI.
-
Interactive visualizations, such as cluster descriptions, spending patterns, gender distribution, and age group analysis, were created to extract key insights.
Dashboard-
- Enhanced visuals with filters, slicers, and drill-down options to make the insights actionable.
HOW TO RUN THIS PROJECT-
- MACHINE LEARNING MODEL-
- Open the python script in Jupyter Notebook or open the Colab file directly.
- Install required libraries.
- Run the notebook to train the K-means clustering model.
- Save the output data with cluster labels.
- POWER BI-
- Import the processed dataset into Power BI.
- Use the .pbix file to open the pre-built dashboard or recreate the visuals using imported data.
- Review interactive dashboard for insights.
KEY INSIGHTS-
- Customer Distribution-
- The majority of customers (81) belong to the "Average" cluster, indicating balanced income and spending patterns.
- High-income and high-spending customers account for 39 individuals, showcasing a key segment for premium products or services.
- The "Low Income, High Spend" cluster (23 customers) could signify customers who prioritize discretionary spending despite limited income.
- Spending Behavior:
- The top-performing cluster by spending score is "High Income, High Spend," with an average score of 82.13, representing the most valuable customer group.
- The "Low Income, High Spend" cluster has a notable spending score of 79.36, which could indicate potential financial strain or a preference for high-value purchases.
- Income-to-Spending Ratio:
- The overall income-to-spending ratio is 1.21, reflecting a moderate balance between income and spending across all clusters.
- Clusters with lower ratios may represent overspending or high-value-focused customer behaviors.
- Gender Distribution:
- Males constitute a slightly higher proportion (56%) compared to females (44%), indicating potential for gender-focused marketing strategies.
- Age Group Analysis:
- Customers aged 30-40 form the largest group (60 customers), followed by the 20-30 age group (45 customers). These age groups might be the primary target audience for marketing campaigns.
- The youngest (<20) and oldest (50-70) groups are the smallest segments, with only 17 customers each, suggesting minimal engagement from these demographics.
- Income and Spending Correlation:
- The scatter plot demonstrates a clear pattern where higher income often correlates with higher spending scores. However, outliers (e.g., low-income, high-spend customers) highlight unique opportunities or risks.