• InvoiceNo: Invoice number. Nominal, a 6-digit integral number uniquely assigned to each transaction. If this code starts with letter 'c', it indicates a cancellation.

• StockCode: Product (item) code. Nominal, a 5-digit integral number uniquely assigned to each distinct product.

• Description: Product (item) name. Nominal.

• Quantity: The quantities of each product (item) per transaction. Numeric.

• InvoiceDate: Invice Date and time. Numeric, the day and time when each transaction was generated.

• UnitPrice: Unit price. Numeric, Product price per unit in sterling.

• CustomerID: Customer number. Nominal, a 5-digit integral number uniquely assigned to each customer.

• Country: Country name. Nominal, the name of the country where each customer resides.

• Targeted Marketing: By segmenting customers based on their behavior, interests and demographics, businesses can tailor their marketing campaigns and promotions to specific groups, resulting in higher response rates and conversions.
• Cross-selling and Upselling: Customer segmentation can help businesses identify customers who are likely to purchase complementary or premium products, and target them with personalized recommendations.
• Customer Retention: By understanding customer behavior and preferences, businesses can design and implement strategies to reduce churn and increase customer loyalty.
• Personalized Customer Experience: Segmentation can help businesses understand customer needs and tailor their offerings, such as product recommendations, customer service, and overall experience, to meet those needs.
• Improved Customer Insights: Customer segmentation provides valuable insights into customer behavior and preferences, which can inform future business decisions and strategies.
• Optimization of Resource Allocation: By prioritizing and focusing resources on high-value customer segments, businesses can optimize their marketing and sales efforts, reduce costs, and increase ROI.

• Implemented libraries such as pandas, numpy,matpplotlib,seaborn,Scikit-learn
• Got the descriptive information of the data.
• checked for null values if any
• Checked for Outliers and duplicated values if any.

• got rid of rows with null customerID
• got rid of duplicated values
• Dropped the id column since it is of no significance to our model training
• Imputed null values with knn imputer and simple imputer
• treated outliers

• visualized maximum ordering countries.
• Visualized month wise orders
• checked the distribution of Revenue with the help of boxplot.

• Applied RFM analysis which allows you to segment customer by the frequency and the value of purchases and how recent was the purchase.
• labelled customer based on the RFM score range 1-4
• Performed log transforamtion on RFM features to noramalize them.

Model training :

• Used Elbow method, silhouete score, davies bouldin score to come up with optimum value of cluster for K-means
• plotted dendogram baased on agglomerative hierrachical clustering.
• found optimum cluster value to be 2 in both the clustering methods.