Fuzzy C-Means for Fraud Detection in Large Transaction Data Sets

Christer Carlsson; Markku Heikkilä; Xiaolu Wang

doi:10.1109/FUZZ-IEEE.2018.8491498

Fuzzy C-Means for Fraud Detection in Large Transaction Data Sets

Christer Carlsson, Markku Heikkilä, Xiaolu Wang

Research output: Chapter in Book/Conference proceeding › Conference contribution › Scientific › peer-review

1 Citation (Scopus)

Abstract

Large, multinational corporations typically have to process hundreds and thousands of vendor invoices every day coming from hundreds or thousands of vendors in different currencies from many countries. This has forced the adoption of simple and fast rules of thumb to accept and pay invoices below a chosen value cutoff without further scrutiny. This has invited fraud, which is hard to detect and react to, as the transaction data sets are very large. Fuzzy C-means has turned out to be effective to identify and sort out transactions, which are similar to typical fraud cases.

Original language	Undefined/Unknown
Title of host publication	2018 IEEE International Conference on Fuzzy Systems (FUZZ)
Publisher	The Institute of Electrical and Electronics Engineers
Pages	–
ISBN (Print)	978-1-5090-6020-7
DOIs	https://doi.org/10.1109/FUZZ-IEEE.2018.8491498
Publication status	Published - 2018
MoE publication type	A4 Article in a conference publication
Event	2018 IEEE International Conference on Fuzzy Systems (FUZZ) - 2018 IEEE International Conference on Fuzzy Systems (FUZZ) Duration: 8 Jul 2018 → 13 Jul 2018

Conference

Conference	2018 IEEE International Conference on Fuzzy Systems (FUZZ)
Period	08/07/18 → 13/07/18

Keywords

Analytics
Fraud detection
Fuzzy C-Means

Access to Document

10.1109/FUZZ-IEEE.2018.8491498

https://ieeexplore.ieee.org/document/8491498

Cite this

@inproceedings{656e30a7620a4057b26d2c3c088f4855,

title = "Fuzzy C-Means for Fraud Detection in Large Transaction Data Sets",

abstract = "Large, multinational corporations typically have to process hundreds and thousands of vendor invoices every day coming from hundreds or thousands of vendors in different currencies from many countries. This has forced the adoption of simple and fast rules of thumb to accept and pay invoices below a chosen value cutoff without further scrutiny. This has invited fraud, which is hard to detect and react to, as the transaction data sets are very large. Fuzzy C-means has turned out to be effective to identify and sort out transactions, which are similar to typical fraud cases.",

keywords = "Analytics, Fraud detection, Fuzzy C-Means, Analytics, Fraud detection, Fuzzy C-Means, Analytics, Fraud detection, Fuzzy C-Means",

author = "Christer Carlsson and Markku Heikkil{\"a} and Xiaolu Wang",

year = "2018",

doi = "10.1109/FUZZ-IEEE.2018.8491498",

language = "Odefinierat/ok{\"a}nt",

isbn = "978-1-5090-6020-7",

pages = "–",

booktitle = "2018 IEEE International Conference on Fuzzy Systems (FUZZ)",

publisher = "The Institute of Electrical and Electronics Engineers",

note = "2018 IEEE International Conference on Fuzzy Systems (FUZZ) ; Conference date: 08-07-2018 Through 13-07-2018",

}

TY - GEN

T1 - Fuzzy C-Means for Fraud Detection in Large Transaction Data Sets

AU - Carlsson, Christer

AU - Heikkilä, Markku

AU - Wang, Xiaolu

PY - 2018

Y1 - 2018

N2 - Large, multinational corporations typically have to process hundreds and thousands of vendor invoices every day coming from hundreds or thousands of vendors in different currencies from many countries. This has forced the adoption of simple and fast rules of thumb to accept and pay invoices below a chosen value cutoff without further scrutiny. This has invited fraud, which is hard to detect and react to, as the transaction data sets are very large. Fuzzy C-means has turned out to be effective to identify and sort out transactions, which are similar to typical fraud cases.

AB - Large, multinational corporations typically have to process hundreds and thousands of vendor invoices every day coming from hundreds or thousands of vendors in different currencies from many countries. This has forced the adoption of simple and fast rules of thumb to accept and pay invoices below a chosen value cutoff without further scrutiny. This has invited fraud, which is hard to detect and react to, as the transaction data sets are very large. Fuzzy C-means has turned out to be effective to identify and sort out transactions, which are similar to typical fraud cases.

KW - Analytics

KW - Fraud detection

KW - Fuzzy C-Means

KW - Analytics

KW - Fraud detection

KW - Fuzzy C-Means

KW - Analytics

KW - Fraud detection

KW - Fuzzy C-Means

U2 - 10.1109/FUZZ-IEEE.2018.8491498

DO - 10.1109/FUZZ-IEEE.2018.8491498

M3 - Konferensbidrag

SN - 978-1-5090-6020-7

SP - –

BT - 2018 IEEE International Conference on Fuzzy Systems (FUZZ)

PB - The Institute of Electrical and Electronics Engineers

T2 - 2018 IEEE International Conference on Fuzzy Systems (FUZZ)

Y2 - 8 July 2018 through 13 July 2018

ER -