Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

International Journal of Electronics and Telecommunications

FT-FM: A Financial-Transaction-Specific Quantum Feature Map for Variational Quantum Classifiers in Fraud Detectio

Jean Marie Vianney Sindayigaya

Published Jul 31, 2026• 6 citations
Publisher

Abstract

Preliminary empirical characterisation on synthetic financial-transaction data confirms FT-FM's predicted expressibility advantage over generic feature maps (Kullback– Leibler divergence to the Haar distribution of 4.01 ± 1.16 versus 21.24 ± 0.79 for a ZZ-like baseline and 19.53 ± 0.43 for an amplitude-like baseline at matched qubit count), and exposes a kernel-concentration phenomenon (Meyer–Wallach Q = 0.948 ± 0.004) that motivates a refined construction (FT-FM-lite) with a tunable entanglement-strength hyperparameter; full empirical benchmarking on Kaggle CCF, CICIDS-2017, and UNSW-NB15 is the subject of a forthcoming paper in this series.

DOI
10.24425/ijet.2026.1729

Related papers

FT-FM-lite: Tunable-Entanglement Refinement and Empirical Benchmarking on Financial-Cybersecurity Datasets

Jean Marie Vianney Sindayigaya

The financial-transaction feature map FT-FM, introduced in the first paper of this series, is empirically more expressive than generic quantum encodings but operates in a near-maximally entangled regime in which quantum-kernel values concentrate and kernel-based classification fails. This paper introduces FT-FM-lite, a refinement of FT-FM with a single entanglement-strength hyperparameter β ∈ [0, 1] that preserves the domain-informed encoding axioms while allowing the feature map to operate in the intermediate expressibility–trainability regime; β = 1 recovers FT-FM exactly. A validation-based protocol selects β per training split. In a ten-seed synthetic diagnostic against exactly implemented matched controls, validation-selected FT-FM-lite attains a mean held-out F1 of 0.620 (95% bootstrap CI [0.530, 0.709]), exceeding exact amplitude encoding (0.362) and Qiskit’s ZZFeatureMap (0.033) with Holm–Bonferroni-corrected significance, while its advantage over full-strength FT-FM (0.490) is not significant after correction. We further specify a leakage-controlled, case-control benchmarking protocol with paired Wilcoxon significance testing and apply it to the Kaggle Credit-Card Fraud dataset over ten repeated stratified case-control splits. In contrast to the synthetic diagnostic, validation selection chose full strength (β = 1) in every seed, and FT-FM-lite attained a mean case-control test F1 of 0.891 (95% bootstrap CI [0.870, 0.907]; ROC-AUC 0.971; PR-AUC 0.976) — significantly above exact amplitude encoding and significantly below a matched RBF-SVM after Holm–Bonferroni correction, and statistically indistinguishable from Qiskit’s ZZFeatureMap and the remaining classical baselines. Balanced-cohort metrics are benchmark values, not deployment-prevalence estimates.

International Journal of Electronics and Telecommunications•Oct 29, 2026•0 citations

Fidelity-Based Quantum Kernels for Hybrid Financial Fraud Detection Pipelines

Fathima Nihla Latheef, Jismina T A, Rubell Marion Lincy G

Financial fraud detection remains difficult due to extreme class imbalance and subtle, nonlinear patterns in transactional data. We investigate a hybrid quantum–classical approach that couples a fidelity-based quantum kernel with a logistic regression classifier (QEKLR). A custom ZZ-inspired feature map comprising Hadamards, mixed-axis Ry/Rzencodings, tunable Ising-ZZ cross terms, and data re-uploading, embeds PCA-reduced, Min–Max scaled features into quantum states. The resulting fidelity kernel is evaluated and used within kernel logistic regression. Experiments on two benchmarks: BankSim (synthetic) and the European Credit Card Fraud dataset, use balanced subsets created via Random Under-Sampling. The proposed model attains the best performance on BankSim (AUC = 0.96; strongest F1), and is competitive on the credit-card dataset (AUC = 0.94), closely tracking XGBoost (AUC = 0.96) while outperforming logistic regression and a variational quantum classifier baseline. These results indicate that quantum kernel methods can provide practical gains on tabular fraud tasks while remaining aligned with current NISQ constraints. We also discuss the quadratic cost of exact kernel evaluation and outline scalable approximations for future deployment on larger, highly imbalanced datasets.

IEEE International Conference on Emerging Computing and Intelligent Technologies (ICoECIT)•Jan 28, 2026•0 citations

Comparative Analysis and Noise Robustness Study of Quantum Kernel Methods and Variational Quantum Classifiers for Financial Fraud Detection

Ionuț-Cosmin Dinuț, Rodica Constantinescu, Bogdan Alexandrescu

Quantum machine learning on near-term noisy quantum devices has generated substantial theoretical interest, but rigorous empirical comparisons under realistic noise on practically relevant data remain scarce. This paper compares two paradigmatic quantum learning models, a Quantum Support Vector Machine (QSVM) built on the ZZFeatureMap quantum kernel and a Variational Quantum Classifier (VQC) with an EfficientSU2/RealAmplitudes ansatz, against tuned classical baselines (SVM with four kernels, Random Forest, XGBoost, LightGBM and CatBoost) on the ULB Credit Card Fraud dataset (284,807 transactions, 0.17% fraud). All models share an identical 4-qubit PCA-reduced feature space, evaluated on the full unbalanced test fold over 15 fits (3 folds × 5 seeds) and reported as mean ± standard deviation with bootstrap confidence intervals, AUPRC as the primary metric. Noise robustness is assessed under depolarizing noise p∈{0,0.001,0.01,0.05}, with ranking preservation measured directly through Spearman ρ and Kendall τ between the noisy and noiseless decision scores rather than read off AUPRC, alongside the per-paradigm computational cost. At four qubits the classical baselines lead (AUPRC 0.60 to 0.74, CatBoost best), above the VQC (0.494) and the QSVM (0.240); the controlled QSVM-versus-RBF–SVM comparison puts the cost of the quantum kernel at about 0.45 AUPRC. Under noise the QSVM keeps its score ranking (ρ=0.998 at p=0.001, 0.906 at p=0.01) and an operational decision threshold (recall 0.87 to 0.89, stable calibration), while the VQC AUPRC peaks non-monotonically at p=0.01 (0.494 rising to 0.654, then easing to 0.569 at p=0.05) even as its ranking decays monotonically (ρ from 0.72 to near zero), so average precision on its own misrepresents how noise affects it. The quantum models do not surpass the tuned classical reference at four qubits; the contribution is methodological: under noise, AUPRC has to be read together with a genuine rank statistic, because the two can move in opposite directions.

Electronics•Jun 5, 2026•1 citation

Where a domain specific quantum feature map pays: a cross-domain

Jean Marie Vianney Sindayigaya

Claims about quantum machine learning are routinely established on a single dataset and generalised informally. This paper tests one such claim directly. Earlier work in this series found, under a leakage-controlled predeclared protocol, that a financial-transaction-specific quantum feature map does not outperform matched classical baselines on the Kaggle Credit-Card Fraud dataset. We freeze source-column contracts and an experimental protocol in advance, inherit every predicted direction from that published result, and replicate on CICIDS-2017 (2.8 million flows) and UNSW-NB15 (257,673 records) — six paired comparisons under one Holm–Bonferroni correction. The replication is partial, and the failure is informative. The map’s advantage over naive amplitude encoding replicates on both datasets and is now established across three datasets and two domains. But the prediction that a matched RBF-SVM would win is significantly reversed on UNSW-NB15, where the quantum kernel wins by a median 0.021 F1 — identifying a regime in which the approach pays. That result is bounded: a random forest on identical cohorts outperforms the quantum kernel on both datasets, so no general claim of quantum advantage is supportable. Two further findings weaken the premise of domain-specific encoding itself. The advantage over generic quantum encodings differs in sign between the two network datasets, and the unanimous entanglement strength selected on financial data does not reproduce on either. We conclude that quantum kernel results reported on one dataset should not be assumed to transfer, in either direction, and that predeclared cross-domain replication is cheap enough to be routine. Balanced-cohort metrics are benchmark values, not deployment-prevalence estimates.

International Journal of Electronics and Telecommunications•Oct 29, 2026•0 citations

Variational Quantum Classifier for Fraudulent Transaction Detection in Synthetic Banking Data: A Noise-Aware Benchmarking Study Using PKTRON

Dr. Zuhair Ahmed

Abstract Financial fraud represents one of the most costly and persistent threats to the global banking sector, with annual losses exceeding $485 billion USD. Machine learning-based fraud detection has become the industry standard, with classical algorithms such as Support Vector Machines (SVM) and Random Forests achieving near-perfect performance on well-structured datasets. As quantum computing hardware matures, the question arises: can variational quantum classifiers (VQCs) approach or exceed classical performance on banking fraud detection, and how does realistic hardware noise affect their viability? In this work, we address this question through a systematic, noise-aware benchmarking study executed on PKTRON v3.7.3, using the PK Falcon 27Q and PK NoisyLab 8Q virtual quantum processors. We generate a 200-sample synthetic banking transaction dataset with six clinically motivated fraud features — transaction amount, time of day, merchant category, transaction velocity, geographic distance, and account age — and evaluate a 4-qubit, 2-layer Variational Quantum Circuit against classical SVM and Random Forest baselines across four experimental phases. Our results reveal that the ideal VQC achieves 88.00% accuracy and 0.6940 Banking Quantum Readiness Score (BQRS), compared to 100.00% and 98.00% for classical SVM and Random Forest respectively. Under noise, amplitude damping at p = 0.10 achieves the highest noisy BQRS of 0.5189, while depolarizing noise is most destructive, collapsing recall and F1 to zero across all tested strengths. Phase damping produces a counter-intuitive pattern of high AUC with zero recall, indicating preserved ranking ability but collapsed decision boundary. We introduce the Banking Quantum Readiness Score (BQRS) — a recall-weighted composite metric (40% recall, 30% F1, 30% AUC-ROC) — as a standardized benchmark for evaluating quantum classifier readiness for banking deployment. Our findings establish the current performance gap between quantum and classical classifiers on fraud detection and define the noise thresholds that must be overcome for practical quantum advantage in financial cybersecurity.

•Jan 1, 2026•0 citations

Feature-Optimized Quantum Classification Model for Financial Transaction Anomalies

S THANGAPRIYA REX KAMALESH

Abstract The exponential growth of digital financial transactions has increased the demand for intelligent and accurate fraud detection systems. This study proposes a Quantum-Enhanced Fraud Detection Framework that integrates classical feature engineering with advanced quantum machine learning techniques. The framework begins with comprehensive data preprocessing, including Mean, Median, and Mode Imputation to handle missing values and Z-Score Standardization along with Min–Max Scaling to normalize the data. The processed features are then transformed into quantum states using the ZZFeatureMap, enabling the representation of complex, non-linear correlations through quantum entanglement and superposition. The Quantum Kernel evaluates the pairwise similarity of quantum-encoded samples, capturing intricate relationships in high-dimensional feature spaces. Finally, a Quantum Support Vector Machine (QSVM) classifier is trained using the quantum kernel matrix to differentiate between fraudulent and legitimate transactions. The proposed method achieved an accuracy of 91%, demonstrating superior performance over conventional approaches by efficiently identifying subtle fraud patterns. The results validate the potential of Quantum Machine Learning (QML) as a powerful paradigm for financial fraud detection, combining classical data preprocessing with quantum computational advantages to achieve higher generalization and robustness.

•Sep 30, 2026•0 citations
Preliminary empirical characterisation on synthetic financial-transaction data confirms FT-FM's predicted expressibility advantage over generic feature maps (Kullback– Leibler divergence to the Haar distribution of 4.01 ± 1.16 versus 21.24 ± 0.79 for a ZZ-like baseline and 19.53 ± 0.43 for an amplitude-like baseline at matched qubit count), and exposes a kernel-concentration phenomenon (Meyer–Wallach Q = 0.948 ± 0.004) that motivates a refined construction (FT-FM-lite) with a tunable entanglement-strength hyperparameter; full empirical benchmarking on Kaggle CCF, CICIDS-2017, and UNSW-NB15 is the subject of a forthcoming paper in this series.
The financial-transaction feature map FT-FM, introduced in the first paper of this series, is empirically more expressive than generic quantum encodings but operates in a near-maximally entangled regime in which quantum-kernel values concentrate and kernel-based classification fails. This paper introduces FT-FM-lite, a refinement of FT-FM with a single entanglement-strength hyperparameter β ∈ [0, 1] that preserves the domain-informed encoding axioms while allowing the feature map to operate in the intermediate expressibility–trainability regime; β = 1 recovers FT-FM exactly. A validation-based protocol selects β per training split. In a ten-seed synthetic diagnostic against exactly implemented matched controls, validation-selected FT-FM-lite attains a mean held-out F1 of 0.620 (95% bootstrap CI [0.530, 0.709]), exceeding exact amplitude encoding (0.362) and Qiskit’s ZZFeatureMap (0.033) with Holm–Bonferroni-corrected significance, while its advantage over full-strength FT-FM (0.490) is not significant after correction. We further specify a leakage-controlled, case-control benchmarking protocol with paired Wilcoxon significance testing and apply it to the Kaggle Credit-Card Fraud dataset over ten repeated stratified case-control splits. In contrast to the synthetic diagnostic, validation selection chose full strength (β = 1) in every seed, and FT-FM-lite attained a mean case-control test F1 of 0.891 (95% bootstrap CI [0.870, 0.907]; ROC-AUC 0.971; PR-AUC 0.976) — significantly above exact amplitude encoding and significantly below a matched RBF-SVM after Holm–Bonferroni correction, and statistically indistinguishable from Qiskit’s ZZFeatureMap and the remaining classical baselines. Balanced-cohort metrics are benchmark values, not deployment-prevalence estimates.
Financial fraud detection remains difficult due to extreme class imbalance and subtle, nonlinear patterns in transactional data. We investigate a hybrid quantum–classical approach that couples a fidelity-based quantum kernel with a logistic regression classifier (QEKLR). A custom ZZ-inspired feature map comprising Hadamards, mixed-axis Ry/Rzencodings, tunable Ising-ZZ cross terms, and data re-uploading, embeds PCA-reduced, Min–Max scaled features into quantum states. The resulting fidelity kernel is evaluated and used within kernel logistic regression. Experiments on two benchmarks: BankSim (synthetic) and the European Credit Card Fraud dataset, use balanced subsets created via Random Under-Sampling. The proposed model attains the best performance on BankSim (AUC = 0.96; strongest F1), and is competitive on the credit-card dataset (AUC = 0.94), closely tracking XGBoost (AUC = 0.96) while outperforming logistic regression and a variational quantum classifier baseline. These results indicate that quantum kernel methods can provide practical gains on tabular fraud tasks while remaining aligned with current NISQ constraints. We also discuss the quadratic cost of exact kernel evaluation and outline scalable approximations for future deployment on larger, highly imbalanced datasets.
Quantum machine learning on near-term noisy quantum devices has generated substantial theoretical interest, but rigorous empirical comparisons under realistic noise on practically relevant data remain scarce. This paper compares two paradigmatic quantum learning models, a Quantum Support Vector Machine (QSVM) built on the ZZFeatureMap quantum kernel and a Variational Quantum Classifier (VQC) with an EfficientSU2/RealAmplitudes ansatz, against tuned classical baselines (SVM with four kernels, Random Forest, XGBoost, LightGBM and CatBoost) on the ULB Credit Card Fraud dataset (284,807 transactions, 0.17% fraud). All models share an identical 4-qubit PCA-reduced feature space, evaluated on the full unbalanced test fold over 15 fits (3 folds × 5 seeds) and reported as mean ± standard deviation with bootstrap confidence intervals, AUPRC as the primary metric. Noise robustness is assessed under depolarizing noise p∈{0,0.001,0.01,0.05}, with ranking preservation measured directly through Spearman ρ and Kendall τ between the noisy and noiseless decision scores rather than read off AUPRC, alongside the per-paradigm computational cost. At four qubits the classical baselines lead (AUPRC 0.60 to 0.74, CatBoost best), above the VQC (0.494) and the QSVM (0.240); the controlled QSVM-versus-RBF–SVM comparison puts the cost of the quantum kernel at about 0.45 AUPRC. Under noise the QSVM keeps its score ranking (ρ=0.998 at p=0.001, 0.906 at p=0.01) and an operational decision threshold (recall 0.87 to 0.89, stable calibration), while the VQC AUPRC peaks non-monotonically at p=0.01 (0.494 rising to 0.654, then easing to 0.569 at p=0.05) even as its ranking decays monotonically (ρ from 0.72 to near zero), so average precision on its own misrepresents how noise affects it. The quantum models do not surpass the tuned classical reference at four qubits; the contribution is methodological: under noise, AUPRC has to be read together with a genuine rank statistic, because the two can move in opposite directions.
Claims about quantum machine learning are routinely established on a single dataset and generalised informally. This paper tests one such claim directly. Earlier work in this series found, under a leakage-controlled predeclared protocol, that a financial-transaction-specific quantum feature map does not outperform matched classical baselines on the Kaggle Credit-Card Fraud dataset. We freeze source-column contracts and an experimental protocol in advance, inherit every predicted direction from that published result, and replicate on CICIDS-2017 (2.8 million flows) and UNSW-NB15 (257,673 records) — six paired comparisons under one Holm–Bonferroni correction. The replication is partial, and the failure is informative. The map’s advantage over naive amplitude encoding replicates on both datasets and is now established across three datasets and two domains. But the prediction that a matched RBF-SVM would win is significantly reversed on UNSW-NB15, where the quantum kernel wins by a median 0.021 F1 — identifying a regime in which the approach pays. That result is bounded: a random forest on identical cohorts outperforms the quantum kernel on both datasets, so no general claim of quantum advantage is supportable. Two further findings weaken the premise of domain-specific encoding itself. The advantage over generic quantum encodings differs in sign between the two network datasets, and the unanimous entanglement strength selected on financial data does not reproduce on either. We conclude that quantum kernel results reported on one dataset should not be assumed to transfer, in either direction, and that predeclared cross-domain replication is cheap enough to be routine. Balanced-cohort metrics are benchmark values, not deployment-prevalence estimates.
Abstract Financial fraud represents one of the most costly and persistent threats to the global banking sector, with annual losses exceeding $485 billion USD. Machine learning-based fraud detection has become the industry standard, with classical algorithms such as Support Vector Machines (SVM) and Random Forests achieving near-perfect performance on well-structured datasets. As quantum computing hardware matures, the question arises: can variational quantum classifiers (VQCs) approach or exceed classical performance on banking fraud detection, and how does realistic hardware noise affect their viability? In this work, we address this question through a systematic, noise-aware benchmarking study executed on PKTRON v3.7.3, using the PK Falcon 27Q and PK NoisyLab 8Q virtual quantum processors. We generate a 200-sample synthetic banking transaction dataset with six clinically motivated fraud features — transaction amount, time of day, merchant category, transaction velocity, geographic distance, and account age — and evaluate a 4-qubit, 2-layer Variational Quantum Circuit against classical SVM and Random Forest baselines across four experimental phases. Our results reveal that the ideal VQC achieves 88.00% accuracy and 0.6940 Banking Quantum Readiness Score (BQRS), compared to 100.00% and 98.00% for classical SVM and Random Forest respectively. Under noise, amplitude damping at p = 0.10 achieves the highest noisy BQRS of 0.5189, while depolarizing noise is most destructive, collapsing recall and F1 to zero across all tested strengths. Phase damping produces a counter-intuitive pattern of high AUC with zero recall, indicating preserved ranking ability but collapsed decision boundary. We introduce the Banking Quantum Readiness Score (BQRS) — a recall-weighted composite metric (40% recall, 30% F1, 30% AUC-ROC) — as a standardized benchmark for evaluating quantum classifier readiness for banking deployment. Our findings establish the current performance gap between quantum and classical classifiers on fraud detection and define the noise thresholds that must be overcome for practical quantum advantage in financial cybersecurity.
Abstract The exponential growth of digital financial transactions has increased the demand for intelligent and accurate fraud detection systems. This study proposes a Quantum-Enhanced Fraud Detection Framework that integrates classical feature engineering with advanced quantum machine learning techniques. The framework begins with comprehensive data preprocessing, including Mean, Median, and Mode Imputation to handle missing values and Z-Score Standardization along with Min–Max Scaling to normalize the data. The processed features are then transformed into quantum states using the ZZFeatureMap, enabling the representation of complex, non-linear correlations through quantum entanglement and superposition. The Quantum Kernel evaluates the pairwise similarity of quantum-encoded samples, capturing intricate relationships in high-dimensional feature spaces. Finally, a Quantum Support Vector Machine (QSVM) classifier is trained using the quantum kernel matrix to differentiate between fraudulent and legitimate transactions. The proposed method achieved an accuracy of 91%, demonstrating superior performance over conventional approaches by efficiently identifying subtle fraud patterns. The results validate the potential of Quantum Machine Learning (QML) as a powerful paradigm for financial fraud detection, combining classical data preprocessing with quantum computational advantages to achieve higher generalization and robustness.