Abstract: Mining the Banking Customer Behavior Using Clustering and Association Rules Methods

Bank databases are multi-dimensional and comprises of

monthly account records and daily transaction records. This study presents a new two-stage frame-work of customer behavior:
K-means algorithm Apriori association rule inducer.

The K-means algorithm was used to identify groups of

customers based on recency, frequency, monetary (RFM) behavioral scoring predicators It also divides customers into three major profitable groups of customers.

Critical elements of K means cluster analysis

The Lloyd iteration for K-means clustering is as follows:
Start: Begin with an initial group of cluster centers (centroids): Step (t): Assign each object (data record x) to the group that has

the closest centroid to generate the improved set of cluster representatives. Update centroid: compute the positions of the centroids. If the value of the centroids didn't change then stop, else go to Step (t).

K-means is built with data from existing customers, which

include variables from account and transaction data tables. CP and RFM are used as predicated variables to classify each customer into clusters.

Conclusion
The existing customers were divided into three profitable

groups of customers according to their shared behavior and characteristics.

Raring User Transactor User Pamper User

Marketers then can infer the profiles of customers in each

group and propose management strategies appropriate to the each group.

Effects of Usage and Name on Perceptions of New Products

Longitudinal changes in perceptions of both new and existing

brands in the same product.

Specifically pair wise similarity judgments are collected before

and after participation in a 6-occasion choice and usage experiment. t class are studied.
Comparisons are made across both the original similarity

judgments and the resulting group-level perceptual spaces.

studies have been conducted to determine (1) if the repositioning

changed the product's perception in the desired manner (or if the new product came to occupy its intended position in the perceptual space) and (2) if this change in perception had the desired effect on preference and choice.

Hypotheses, Experiment, and Analytic Method

H1: The perceptions of the old brands will change less over the

course of the experiment than the perceptions of the new brands.

H2: The perceptions of the new brands will change: (a) more for

people receiving non descriptive names than those receiving descriptive names, (b) more for people acquiring more pieces of information during the experiment, and (c) more for the people who consumed a greater number of different brands during the experiment.
120 students and staff members participated in a longitudinal

information acquisi-tion and purchase experiment. The subjects were required to choose one of five types of bread each week for six weeks.

Conclusions
Results strongly support hypotheses 1 and 2. Subjects were significantly more consistent in their similarity

judgments of the old brands than the new brands, supporting H1.
Correlations over all 45 pair wise judgments are significantly

higher for the group receiving descriptive names, supporting H2.

The number of brands chosen is positively related to those

correlations among old brands only.