ADS Capstone Chronicles Revised

12

1

(Cooking)

(Memoir)

0.001107

0.082192

8.568493

8

(iPhone)

(Sony Xperia) 0.001291

0.092105

5.944079

9

(Sony Xperia) (iPhone)

0.001291

0.083333

5.944079

2

(Science)

(Cooking)

0.001107

0.075

5.569521

3

(Cooking)

(Science)

0.001107

0.082192

5.569521

6

(History)

(Psychology)

0.001107

0.061856

5.00477

7

(Psychology)

(History)

0.001107

0.089552

5.00477

Slate. The lift value is 11.733766, suggesting that transactions containing Huawei MediaPad are approximately 11.73 times more likely to also contain Google Pixel Slate compared to transactions in general. loyal, spender, hibernating Champions customers are high in recency, frequency and monetary. These customers are active, spending a lot on each purchasing time. Potential Loyalists are those customers who have purchased recently, and their purchase frequency is high. If they can be pushed by better product recommendations, they can be Champion customers. Big Spenders are customers with high monetary scores, known for significant spending when they make purchases. Hibernating customers have low recency scores, having not made purchases recently. At risk customers are low in recency, frequency and monetary. These customers are inactive and for a long time, they are not purchasing anything. Lastly, the category of Others includes customers who do not fit specific behavioral profiles. These segments help businesses tailor their strategies to effectively engage and retain different types of customers. Figure 5 Bar Chart Visualizing Distribution of Customers Across Different RFM (Recency, Frequency, Monetary) Segments

Customer segmentation based on purchasing behavior, various categories emerge. At-risk customers are characterized by low scores in all three dimensions: 1) no recent purchases, 2) buying infrequently, and 3) exhibiting minimal spending. This category reflects customers who are less engaged with the brand and may require targeted interventions to re-engage them. Loyal Customers, in contrast, display high recency scores, characterized by recent purchases and strong engagement with the brand. Frequent transactions and higher spending levels suggest a robust relationship with the brand. The analysis of association rules uncovers valuable insights into customer purchasing behaviors and product relationships. An example of association rules is rule 4: If a customer purchases a Huawei MediaPad, there is a notable association with purchasing a Google Pixel Slate. This rule demonstrates a significant purchasing pattern where the acquisition of one product is strongly linked to the purchase of the other. Rule 4: (Huawei MediaPad) → (Google Pixel Slate) is supported 0.001107, indicating that 0.1107% of all transactions contain both Huawei MediaPad and Google Pixel Slate. The confidence level is 0.142857, meaning that among all transactions containing Huawei MediaPad, 14.29% also include Google Pixel

56