Instance-level explanations for fraud detection

D.A.C. Collaris, Leo M. Vink, J.J. van Wijk

Research output: Contribution to conference › Poster › Academic

14 Downloads (Pure)

Abstract

Fraud detection is a difficult problem that can benefit from predictive modeling. However, the verification of a prediction is challenging; for a single insurance policy, the model only provides a prediction score. We present a case study where we reflect on different instance-level model explanation techniques to aid a fraud detection team in their work. To this end, we designed two novel dashboards combining various state-of-the-art explanation techniques. These enable the domain expert to analyze and understand predictions, dramatically speeding up the process of filtering potential fraud cases.
Original languageEnglish
Number of pages1
Publication statusPublished - 27 Nov 2018
Event5th Data Science Summit (DSSE 2018) - Muziekgebouw Frits Philips, Eindhoven, Netherlands
Duration: 27 Nov 2018 → 27 Nov 2018
Conference number: 5

Other

Other5th Data Science Summit (DSSE 2018)
Abbreviated titleDSSE 2018
CountryNetherlands
CityEindhoven
Period27/11/18 → 27/11/18

Fingerprint

Insurance

Keywords

  • Interpretability
  • Explanation
  • Machine learning
  • Sensitivity analysis
  • Local rule extraction
  • Instance-level explanations
  • Fraud detection
  • Case study

Cite this

Collaris, D. A. C., Vink, L. M., & van Wijk, J. J. (2018). Instance-level explanations for fraud detection. Poster session presented at 5th Data Science Summit (DSSE 2018), Eindhoven, Netherlands.
Collaris, D.A.C. ; Vink, Leo M. ; van Wijk, J.J. / Instance-level explanations for fraud detection. Poster session presented at 5th Data Science Summit (DSSE 2018), Eindhoven, Netherlands.1 p.
@conference{bac8fecb17d2484897f310ac80fb3dbc,
title = "Instance-level explanations for fraud detection",
abstract = "Fraud detection is a difficult problem that can benefit from predictive modeling. However, the verification of a prediction is challenging; for a single insurance policy, the model only provides a prediction score. We present a case study where we reflect on different instance-level model explanation techniques to aid a fraud detection team in their work. To this end, we designed two novel dashboards combining various state-of-the-art explanation techniques. These enable the domain expert to analyze and understand predictions, dramatically speeding up the process of filtering potential fraud cases.",
keywords = "Interpretability, Explanation, Machine learning, Sensitivity analysis, Local rule extraction, Instance-level explanations, Fraud detection, Case study",
author = "D.A.C. Collaris and Vink, {Leo M.} and {van Wijk}, J.J.",
year = "2018",
month = "11",
day = "27",
language = "English",
note = "5th Data Science Summit (DSSE 2018), DSSE 2018 ; Conference date: 27-11-2018 Through 27-11-2018",

}

Collaris, DAC, Vink, LM & van Wijk, JJ 2018, 'Instance-level explanations for fraud detection' 5th Data Science Summit (DSSE 2018), Eindhoven, Netherlands, 27/11/18 - 27/11/18, .

Instance-level explanations for fraud detection. / Collaris, D.A.C.; Vink, Leo M.; van Wijk, J.J.

2018. Poster session presented at 5th Data Science Summit (DSSE 2018), Eindhoven, Netherlands.

Research output: Contribution to conference › Poster › Academic

TY - CONF

T1 - Instance-level explanations for fraud detection

AU - Collaris, D.A.C.

AU - Vink, Leo M.

AU - van Wijk, J.J.

PY - 2018/11/27

Y1 - 2018/11/27

N2 - Fraud detection is a difficult problem that can benefit from predictive modeling. However, the verification of a prediction is challenging; for a single insurance policy, the model only provides a prediction score. We present a case study where we reflect on different instance-level model explanation techniques to aid a fraud detection team in their work. To this end, we designed two novel dashboards combining various state-of-the-art explanation techniques. These enable the domain expert to analyze and understand predictions, dramatically speeding up the process of filtering potential fraud cases.

AB - Fraud detection is a difficult problem that can benefit from predictive modeling. However, the verification of a prediction is challenging; for a single insurance policy, the model only provides a prediction score. We present a case study where we reflect on different instance-level model explanation techniques to aid a fraud detection team in their work. To this end, we designed two novel dashboards combining various state-of-the-art explanation techniques. These enable the domain expert to analyze and understand predictions, dramatically speeding up the process of filtering potential fraud cases.

KW - Interpretability

KW - Explanation

KW - Machine learning

KW - Sensitivity analysis

KW - Local rule extraction

KW - Instance-level explanations

KW - Fraud detection

KW - Case study

M3 - Poster

ER -

Collaris DAC, Vink LM, van Wijk JJ. Instance-level explanations for fraud detection. 2018. Poster session presented at 5th Data Science Summit (DSSE 2018), Eindhoven, Netherlands.

Access to Document

    Related content

    Research Output

    Instance-level explanations for fraud detection: a case study

    Research output: Contribution to conference › Paper › Academic