Credit card fraud detection remains a significant challenge in the financial industry, necessitating advanced models to identify fraudulent activities while minimizing false positives accurately. Traditional machine learning approaches, such as Multilayer Perceptrons (MLP), have been widely used but often struggle with interpretability and parameter optimization issues. Kolmogorov-Arnold Networks (KAN) present a promising alternative by addressing these limitations through their inherent structure, which allows for more interpretable and potentially more accurate models. This paper explores the application of KAN in the context of credit card fraud detection, motivated by the need for more effective and interpretable solutions. We implement and evaluate three MLP, KAN, and efficient KAN models using two publicly available credit card fraud datasets. Our experimental results demonstrate that both KAN and efficient KAN significantly outperform the traditional MLP model in terms of detection accuracy while reducing the number of parameters compared to MLP. The findings underscore the potential of KAN and its efficient variant as superior alternatives for credit card fraud detection, offering enhanced accuracy and interpretability. This study provides valuable insights into model performance and parameter efficiency, guiding future research and practical applications in fraud detection systems.

This paper presents a comparative analysis of several decision models for detecting Structured Query Language (SQL) injection attacks, which remain one of the most prevalent and serious security threats to web applications. SQL injection enables attackers to exploit databases, gain unauthorized access, and manipulate data. Traditional detection methods often struggle due to the constantly evolving nature of these attacks, the increasing complexity of modern web applications, and the lack of transparency in the decision-making processes of machine learning models. To address these challenges, we evaluated the performance of various models, including decision tree, random forest, XGBoost, AdaBoost, Gradient Boosting Decision Tree (GBDT), and Histogram Gradient Boosting Decision Tree (HGBDT), using a comprehensive SQL injection dataset. The primary motivation behind our approach is to leverage the strengths of ensemble learning and boosting techniques to enhance detection accuracy and robustness against SQL injection attacks. By systematically comparing these models, we aim to identify the most effective algorithms for SQL injection detection systems. Our experiments show that decision tree, random forest, and AdaBoost achieved the highest performance, with an accuracy of 99.50% and an F1 score of 99.33%. Additionally, we applied SHapley Additive exPlanations (SHAPs) and Local Interpretable Model-agnostic Explanations (LIMEs) for local explainability, illustrating how each model classifies normal and attack cases. This transparency enhances the trustworthiness of our approach to detecting SQL injection attacks. These findings highlight the potential of ensemble methods to provide reliable and efficient solutions for detecting SQL injection attacks, thereby improving the security of web applications.

As Controller Area Network (CAN) Intrusion Detection Systems (IDS) become increasingly sophisticated and incorporate deep learning architectures, adversarial attacks have evolved correspondingly over the past decade. To assess the vulnerability of deep learning-based CAN IDS to evasion attacks, we developed a Double Deep Q-Network (DDQN) agent capable of performing adversarial attacks against two critical attack types: Denial of Service (DoS) and Fuzzing attacks. Our methodology employs reinforcement learning to systematically discover perturbation strategies that can evade detection while maintaining attack functionality. We evaluated our approach across eight different IDS architectures, including RNN variants (RNN, LSTM, GRU) and Transformer-based models, using augmented datasets collected from two distinct vehicle environments (2016 Chevrolet Silverado and 2017 Subaru Forester). To enhance dataset diversity and address class imbalance, we applied data augmentation techniques including CaringCaribou-based attack generation and SMOTE-ENN hybrid sampling. Experimental results demonstrate remarkable attack effectiveness across different architectures and vehicle platforms. The DDQN agent achieved a perfect 100% evasion success rate against all RNN-variant IDS models when tested with fuzzing attacks on the training vehicle dataset, demonstrating strong transferability with an average 98.6% success rate on a different vehicle dataset. For DoS attacks, the agent exhibited variable performance with an overall average success rate of 49.32%, revealing dataset-specific vulnerabilities where success rates ranged from moderate performance on one vehicle dataset to near-perfect evasion on another. Notably, Transformer-based IDS demonstrated significantly greater robustness against both attack types compared to RNN variants, with success rates dropping to 66.77% for Fuzzing attacks on one dataset and complete resistance (0% success rate) on another. These findings highlight the importance of architectural diversity in building resilient IDS systems and underscore the persistent vulnerabilities of current deep learning approaches to sophisticated adversarial attacks. Our work contributes to the understanding of adversarial threats in automotive cybersecurity and emphasizes the need for adversarial-aware defense mechanisms in critical vehicle network infrastructures.

The Controller Area Network (CAN) is crucial for automotive safety, yet remains vulnerable to various fuzzing attacks that can compromise vehicle operations. This paper presents a comprehensive detection framework that identifies both common CAN vulnerabilities (DoS, Spoofing, Replay, and general Fuzzing) and specific fuzzer attack types (identity, replay, random, brute force, and mutation-based) using deep learning-based models. We evaluate four recurrent neural network architectures, including standard RNN, Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and Gated Recurrent Unit (GRU), across three CAN datasets: the publicly available Hacking Car and OTIDS datasets, along with our proprietary Multi-FuzzerCAN dataset. To address the inherent class imbalance in attack data, we implement targeted random oversampling and random undersampling techniques tailored to each dataset’s distribution characteristics, significantly improving model performance. Our experimental results demonstrate that all four models achieve exceptional accuracy, recall, precision, and F1 scores across all datasets, with LSTM and BiLSTM architectures consistently outperforming other approaches. These findings establish the viability of combining deep learning models with data balancing techniques to enhance the resilience of automotive CAN networks against diverse cyber threats, especially multi-fuzzer attacks.

Text recognition on coffee bean package labels is of great importance for product tracking and brand verification, but it poses a challenge due to variations in image quality, packaging materials, and environmental conditions. In this paper, we propose a pipeline that combines several image enhancement techniques and is followed by an Optical Character Recognition (OCR) model based on vision–language (VL) Qwen VL variants, conditioned by structured prompts. To facilitate the evaluation, we construct a coffee bean package image set containing two subsets, namely low-resolution (LRCB) and high-resolution coffee bean image sets (HRCB), enclosing multiple real-world challenges. These cases involve various packaging types (bottles and bags), label sides (front and back), rotation, and different illumination. To address the image quality problem, we design a dedicated preprocessing pipeline for package label situations. We develop and evaluate four Qwen-VL OCR variants with prompt engineering, which are compared against four baselines: DocTR, PaddleOCR, EasyOCR, and Tesseract. Extensive comparison using various metrics, including the Levenshtein distance, Cosine similarity, Jaccard index, Exact Match, BLEU score, and ROUGE scores (ROUGE-1, ROUGE-2, and ROUGE-L), proves significant improvements upon the baselines. In addition, the public POIE dataset validation test proves how well the framework can generalize, thus demonstrating its practicality and reliability for label recognition.

This paper presents a comparative analysis of several decision models for detecting Structured Query Language (SQL) injection attacks, which remain one of the most prevalent and serious security threats to web applications. SQL injection enables attackers to exploit databases, gaining unauthorized access, and manipulating data. Traditional detection methods often struggle due to the constantly evolving nature of these attacks, the increasing complexity of modern web applications, and the lack of transparency in the decision-making processes of machine learning models. To address these challenges, we evaluated the performance of various models, including Decision Tree, Random Forest, XGBoost, AdaBoost, Gradient Boosting Decision Tree (GBDT), and Histogram Gradient Boosting Decision Tree (HGBDT), using a comprehensive SQL injection dataset. The primary motivation behind our approach is to leverage the strengths of ensemble learning and boosting techniques to enhance detection accuracy and robustness against SQL injection attacks. By systematically comparing these models, we aim to identify the most effective algorithms for SQL injection detection systems. Our experiments show that Decision Tree, Random Forest, and AdaBoost achieved the highest performance, with an accuracy of 99.50% and an F1 score of 99.33%. Additionally, we applied SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) for local explainability, illustrating how each model classifies normal and attack cases. This transparency enhances the trustworthiness of our approach in detecting SQL injection attacks. These findings highlight the potential of ensemble methods to provide reliable and efficient solutions for detecting SQL injection attacks, thereby improving the security of web applications.