Selected work

Representative Papers

Performance Benchmarking and Optimisation of Clustering Algorithms for Local and Non-Local Similarity Measure in Medical Image Analysis

Jul 10, 2026

This study addresses the challenge in medical image post-processing of simultaneously preserving local details and exploiting non-local self-similarity—a balance that traditional global low-rank methods struggle to achieve. The authors present the first systematic evaluation and optimization of five clustering algorithms—k-means, mini-batch k-means, agglomerative hierarchical clustering, BIRCH, and bisecting k-means—across multimodal medical images, including MRI, ultrasound, and chest X-rays. Algorithm performance is assessed using Silhouette, Davies–Bouldin (DB), and Calinski–Harabasz (CH) indices, with hyperparameters tuned via random search. Results reveal that agglomerative clustering achieves the best performance on MRI and ultrasound, while mini-batch k-means offers the most balanced results for X-ray images. Standard k-means and bisecting k-means exhibit high inter-cluster separation but large intra-cluster variation, whereas BIRCH underperforms overall, highlighting a fundamental trade-off between clustering efficiency and the preservation of local image details.

0 citationsRead paper

An extendable, integrated, and dynamic approach to forecasting and stress-testing credit risk

Jun 17, 2026

This study addresses the limitations of traditional stress testing approaches, which often decouple loan origination dynamics and neglect the correlation structure among risk indicators, thereby failing to capture the evolving nature of credit risk. To overcome these shortcomings, the paper proposes an integrated, scalable dynamic stress testing framework that, for the first time, jointly models loan issuance and credit risk prediction. The framework employs a multi-state probabilistic model to simulate loan cash flows and embeds macroeconomic stress scenarios directly within Monte Carlo simulations. It allows risk parameters to adjust dynamically in response to both macroeconomic and microeconomic variables and explicitly incorporates the interdependencies among key risk metrics. This approach significantly enhances the realism, flexibility, and forward-looking capability of stress testing, enabling dynamic forecasts of portfolio-level default and loss rates under a wide range of scenarios.

0 citationsRead paper

An Optimised Greedy-Weighted Ensemble Framework for Financial Loan Default Prediction

Mar 19, 2026

This work proposes BlendNet, a particle swarm optimization (PSO)-based weighted ensemble framework designed to address the limitations of traditional models in financial loan default prediction, which stem from nonlinear relationships, class imbalance, and dynamic shifts in borrower behavior. BlendNet integrates tree-based models and neural networks, employing recursive feature elimination for feature selection and a dynamic greedy weighting mechanism that assigns base model weights based on empirical performance. To capture higher-order interactions among model outputs while ensuring both predictive accuracy and well-calibrated probabilities, a neural network meta-learner is introduced in a stacking architecture. Evaluated on the Lending Club dataset, BlendNet achieves an AUC of 0.80, a macro-averaged F1-score of 0.73, and a default recall of 0.81, significantly outperforming individual baseline models.

0 citationsRead paper

Clustering-based Low-Rank Matrix Approximation: An Adaptive Theoretical Analysis with Application to Data Compression

May 13, 2025

Global low-rank matrix approximation (LoRMA) neglects local structural variations, leading to critical detail loss in medical imaging. To address this, we propose an adaptive LoRMA method: overlapping image patches are clustered via k-means based on similarity, and singular value decomposition (SVD) is applied independently to each cluster—preserving global redundancy suppression while enabling local-aware compression. We establish, for the first time, a theoretical framework for adaptive low-rank approximation under overlapping patch clustering, quantifying the trade-off between patch size, compression ratio, and computational cost. This enables joint optimization: high-fidelity reconstruction in clinically critical regions and aggressive compression elsewhere. Evaluated on MRI, ultrasound, CT, and X-ray datasets, our method outperforms global SVD by +1.8 dB PSNR, +0.04 SSIM, −32% MSE, +6.5% IoU, and +11.2% EPI, while significantly suppressing block artifacts and enhancing diagnostic fidelity in lesion regions.

0 citationsRead paper

Duality theory and representations for distributive quasi relation algebras and DInFL-algebras

May 12, 2025

This paper addresses the lack of a duality theory for distributive quasi-relation algebras (DQRA) and involutive FL-algebras, as well as the undecidability of representability. It establishes a systematic duality framework: first, a categorical duality between DQRA and dual involutive FL-algebras (DInFL); second, an ordered relational structure—specifically, a partially ordered frame—for completely perfect algebras, and introduces the novel *bi-pointed Priestley topological frame*, enabling duality extension from completely perfect to arbitrary algebras; third, a complete classification of representability for all algebras of order ≤6. Key contributions include: the first full order-theoretic duality characterizations for both classes; proofs that several algebras are representable as term subreducts of representable relation algebras; and a full representability classification for all algebras up to order six—resolving critical gaps in duality theory and finite representability for relation algebras in algebraic logic.

0 citationsRead paper

Recent publications

Latest Papers

Performance Benchmarking and Optimisation of Clustering Algorithms for Local and Non-Local Similarity Measure in Medical Image Analysis

Jul 10, 2026

This study addresses the challenge in medical image post-processing of simultaneously preserving local details and exploiting non-local self-similarity—a balance that traditional global low-rank methods struggle to achieve. The authors present the first systematic evaluation and optimization of five clustering algorithms—k-means, mini-batch k-means, agglomerative hierarchical clustering, BIRCH, and bisecting k-means—across multimodal medical images, including MRI, ultrasound, and chest X-rays. Algorithm performance is assessed using Silhouette, Davies–Bouldin (DB), and Calinski–Harabasz (CH) indices, with hyperparameters tuned via random search. Results reveal that agglomerative clustering achieves the best performance on MRI and ultrasound, while mini-batch k-means offers the most balanced results for X-ray images. Standard k-means and bisecting k-means exhibit high inter-cluster separation but large intra-cluster variation, whereas BIRCH underperforms overall, highlighting a fundamental trade-off between clustering efficiency and the preservation of local image details.

0 citationsRead paper

An extendable, integrated, and dynamic approach to forecasting and stress-testing credit risk

Jun 17, 2026

This study addresses the limitations of traditional stress testing approaches, which often decouple loan origination dynamics and neglect the correlation structure among risk indicators, thereby failing to capture the evolving nature of credit risk. To overcome these shortcomings, the paper proposes an integrated, scalable dynamic stress testing framework that, for the first time, jointly models loan issuance and credit risk prediction. The framework employs a multi-state probabilistic model to simulate loan cash flows and embeds macroeconomic stress scenarios directly within Monte Carlo simulations. It allows risk parameters to adjust dynamically in response to both macroeconomic and microeconomic variables and explicitly incorporates the interdependencies among key risk metrics. This approach significantly enhances the realism, flexibility, and forward-looking capability of stress testing, enabling dynamic forecasts of portfolio-level default and loss rates under a wide range of scenarios.

0 citationsRead paper

An Optimised Greedy-Weighted Ensemble Framework for Financial Loan Default Prediction

Mar 19, 2026

This work proposes BlendNet, a particle swarm optimization (PSO)-based weighted ensemble framework designed to address the limitations of traditional models in financial loan default prediction, which stem from nonlinear relationships, class imbalance, and dynamic shifts in borrower behavior. BlendNet integrates tree-based models and neural networks, employing recursive feature elimination for feature selection and a dynamic greedy weighting mechanism that assigns base model weights based on empirical performance. To capture higher-order interactions among model outputs while ensuring both predictive accuracy and well-calibrated probabilities, a neural network meta-learner is introduced in a stacking architecture. Evaluated on the Lending Club dataset, BlendNet achieves an AUC of 0.80, a macro-averaged F1-score of 0.73, and a default recall of 0.81, significantly outperforming individual baseline models.

0 citationsRead paper

Clustering-based Low-Rank Matrix Approximation: An Adaptive Theoretical Analysis with Application to Data Compression

May 13, 2025

Global low-rank matrix approximation (LoRMA) neglects local structural variations, leading to critical detail loss in medical imaging. To address this, we propose an adaptive LoRMA method: overlapping image patches are clustered via k-means based on similarity, and singular value decomposition (SVD) is applied independently to each cluster—preserving global redundancy suppression while enabling local-aware compression. We establish, for the first time, a theoretical framework for adaptive low-rank approximation under overlapping patch clustering, quantifying the trade-off between patch size, compression ratio, and computational cost. This enables joint optimization: high-fidelity reconstruction in clinically critical regions and aggressive compression elsewhere. Evaluated on MRI, ultrasound, CT, and X-ray datasets, our method outperforms global SVD by +1.8 dB PSNR, +0.04 SSIM, −32% MSE, +6.5% IoU, and +11.2% EPI, while significantly suppressing block artifacts and enhancing diagnostic fidelity in lesion regions.

0 citationsRead paper

Duality theory and representations for distributive quasi relation algebras and DInFL-algebras

May 12, 2025

This paper addresses the lack of a duality theory for distributive quasi-relation algebras (DQRA) and involutive FL-algebras, as well as the undecidability of representability. It establishes a systematic duality framework: first, a categorical duality between DQRA and dual involutive FL-algebras (DInFL); second, an ordered relational structure—specifically, a partially ordered frame—for completely perfect algebras, and introduces the novel *bi-pointed Priestley topological frame*, enabling duality extension from completely perfect to arbitrary algebras; third, a complete classification of representability for all algebras of order ≤6. Key contributions include: the first full order-theoretic duality characterizations for both classes; proofs that several algebras are representable as term subreducts of representable relation algebras; and a full representability classification for all algebras up to order six—resolving critical gaps in duality theory and finite representability for relation algebras in algebraic logic.

0 citationsRead paper