MétaCan
Menu
Back to cohort
Record W4402891761 · doi:10.1109/access.2024.3468993

Machine Learning Based Method for Insurance Fraud Detection on Class Imbalance Datasets With Missing Values

2024· article· en· W4402891761 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

fundA Canadian funder is recorded on the work.
no affNo Canadian affiliation: this work is invisible to an affiliation-only frame.
No Canadian affiliation. An affiliation-only frame, the usual design, would never have seen this work. It is one of the works that make the case for inverting the frame.

Bibliographic record

VenueIEEE Access · 2024
Typearticle
Languageen
FieldComputer Science
TopicImbalanced Data Classification Techniques
Canadian institutionsnot available
FundersOntario Ministry of Research and InnovationNational Natural Science Foundation of ChinaPrince Sattam bin Abdulaziz University
KeywordsComputer scienceClass (philosophy)Machine learningArtificial intelligenceData mining

Abstract

fetched live from OpenAlex

Insurance fraud is a prevalent issue that insurance companies must face, particularly in the realm of automobile insurance. This type of fraud has significant cost implications for insurance firms and can have a long-term impact on pricing strategies and insurance rates. As a result, accurately predicting and detecting insurance fraud has become a crucial challenge for insurers. The fraud datasets are usually imbalanced, as the number of fraudulent instances is much less than the ligament instances and contains missing values. Prior research has employed machine learning methods to address this class imbalance dataset problem, but there is limited effort handling the class imbalance dataset present in insurance fraud datasets. Moreover, we could not find an overfitting analysis for the relevant predictive models. This paper addresses these two limitations by employing two car insurance company datasets, namely, an Egyptian real-life dataset and a standard dataset. We proposed addressing the missing data and the class imbalance problems with different methods. Then, the predictive models were trained on processed datasets to predict insurance fraud as a classification problem. The classifiers are evaluated on several evaluation metrics. Moreover, we proposed the first overfitting analysis for insurance fraud classifiers, to our knowledge. The obtained results outline that addressing the class imbalance in the insurance fraud detection dataset has a significant positive effect on the performance of the predictive model, while addressing the problem of missing values has a slight effect. Moreover, the proposed methods outperform all of the existing methods on the accuracy metric.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Other design · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: none
Teacher disagreement score0.909
Threshold uncertainty score0.936

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.001
Science and technology studies0.0000.000
Scholarly communication0.0010.002
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.035
GPT teacher head0.353
Teacher spread0.318 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it