C4.5 And C5.0 Algorithm Analysis

A. C4.5 and C5.0 Algorithm:

C4.5 is the decision tree algorithm which has most widely spread usage. It is a developed version of ID3 decision tree algorithm. The C5.0 algorithm (boosting algorithm) is the successive developed state of C4.5 and it has special use for large data sets. To increase the precision and correctness of the C5.0 algorithm, the boosting algorithms are applied in it. It is more rapid and efficient as that of C4.5 and it makes use of memory in more productive way and helps to have smoother trees.

B. CART algorithm:

CART is the classification algorithm which normally uses data in both numerical and nominal form, as the input and also as predicted variables. It has a distinctive dual form which is divided into a structure.

These rules describe the decisions that eventually lead to the prediction.

It uses ID3, C4.5 and C5.0 algorithms, which help to gain percentage accuracy in performance In banking context, Decision tress can be used for Credit card Approval and to detect and prevent Financial frauds with an increased accuracy and performance.

Financial organizations can use these approaches to make their financial services and transactions secure and less vulnerable to fraudulent activities.

K-Means Clustering

(Unsupervised) It is a distance-based algorithm based on clustering that divides the data into a pre-scheduled number of groups/clusters. Each cluster has a unique centroid. Each data element that is in a particular cluster is close to the centroid. K-mean algorithm can be used to make clusters on the basis of the geographic location or job nature or any other feature to offer services accordingly.

This helps to get increased customer retention and help to have an effective customer relationship

By using Data mining techniques and methods, banks can be facilitated in Marketing, Risk Management, Fraud detection and prevention, and customer increased retention. The extraction of rules between different services offered by banks is therefore of great significance for the effective business strategies incorporated by banks. The paper analyses different Decision tree algorithms and k-Means algorithm in context to banking sector. It is concluded that Decision trees help to detect and prevent fraudulent activities while K-mean algorithm can be employed to get increased customer retention and help to have an effective customer relationship