Proceedings of the International Conference on Mathematics, Geometry, Statistics, and Computation (IC-MaGeStiC 2021)

# Classification of Bank Deposit Using Naïve Bayes Classifier (NBC) and K–Nearest Neighbor (K-NN)

Authors
Departement of Mathematics, Faculty of Mathematics and Natural Science, University of Jember, Indonesia.
*Corresponding author. Email: dian_a.fmipa@unej.ac.id
Corresponding Author
Dian Anggraeni
Available Online 8 February 2022.
DOI
10.2991/acsr.k.220202.031How to use a DOI?
Keywords
Classification; Naive Bayes classifier; K-nearest neighbor; Importance variables
Abstract

Banks are financial institutions whose activities are to collect funds from the public in the form of deposits (saving deposit, demand deposit, and time deposit) and distribute them to the public in the form of credit or other forms. Deposits are an alternative for customers because the interest offered on deposits is higher than regular savings. Naïve Bayes Classification (NBC) is a statistical classification method based on Bayes’ theorem that can be used to predict the probability of membership of a class. K-Nearest Neighbor (K-NN) is a method for classifying objects based on the learning data that is closest to the object. This study will use bank customer data consisting of 4521 records and 17 variables. The data is divided into 3 types of training-testing processes, namely 70%:30%, 75%:25%, and 80%:20% and the Ќ -fold cross validation method is used with a value of Ќ =10. The results of this study indicate that the K-NN method is better than the NBC method. Where the best classification of the K-NN method is in the training-testing process of 70%:30% which has an accuracy rate of 89.23%. While the best classification of the NBC method is in the training-testing process of 80%:20% which has an accuracy rate of 84.51%. the K-Nearest Neighbor and the Naïve Bayes Classifier method show the same results on the importance variables. Where out of 16 variables in classifying banking customers, of which 5 which have the most influence are the duration of time the bank contacted its customers, the results of the previous deposit offer, the last month contacted the customer, the type of communication used by the customer, and the number of contacts the bank had made prior to the promotion of opening a deposit.

Open Access
This is an open access article under the CC BY-NC license.

Volume Title
Proceedings of the International Conference on Mathematics, Geometry, Statistics, and Computation (IC-MaGeStiC 2021)
Series
Publication Date
8 February 2022
ISBN
10.2991/acsr.k.220202.031
ISSN
2352-538X
DOI
10.2991/acsr.k.220202.031How to use a DOI?
Open Access
This is an open access article under the CC BY-NC license.

TY  - CONF
AU  - Dian Anggraeni
AU  - Yuliani Setia Dewi