Unsupervised Anomaly Detection for SIM Card Fraud Using Autoencoder-Based Reconstruction Error Analysis Zakaria Toumi, Kheyreddine Djouzi, Mohamed Islam Rahil Proceedings of the 2025 International Conference on Artificial Intelligence and Innovative Applications Aiia 2025, 2025 Detecting commercial fraud in telecommunications networks remains critically challenging as sophisticated schemes systematically evade traditional rule-based detection systems. This paper introduces an unsupervised deep learning framework for SIM card fraud detection, rigorously evaluated on a real-world dataset comprising 1.5 million commercial transactions. From eight raw transactional attributes, we engineer a discriminative feature space of 26 behavioral and temporal indicators. We systematically evaluate multiple autoencoder architectures "shallow, medium, deep, and denoising variants" and a Variational Autoencoder (VAE), benchmarking them against a One-Class Support Vector Machine (OCSVM) baseline. Evaluation on a labeled test set of 291000 transactions reveals that the Deep Autoencoder achieves superior performance with an Area Under the Receiver Operating Characteristic Curve (AUC-ROC) of 0.885, substantially outperforming all competing approaches. Beyond raw detection accuracy, our analysis establishes that the model excels at riskbased transaction ranking, providing an effective prioritization mechanism for human investigators. We integrate t-distributed Stochastic Neighbor Embedding (t-SNE) for latent space visualization, SHapley Additive exPlanations (SHAP) for model interpretability, and Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN) for anomaly characterization, uncovering over 4000 distinct anomaly typologies. This framework delivers not merely alerts, but actionable, interpretable intelligence into the operational patterns underlying fraudulent activities.
Collisions-Resistant Hash Function Based on a Logistics Map Nassim Zaouia, Kheyreddine Djouzi, Mehammed Daoui Proceedings of the Workshop on Enabling Technologies Infrastructure for Collaborative Enterprises Wetice, 2023 Hash functions play a pivotal role in the field of cryptography, serving as a fundamental technique for ensuring data security. Moreover, cryptographic hash functions can be utilized to maintain the integrity of expansive datasets, including those stored on hard drives or encompassing financial data. In this context, chaotic maps, renowned for their unpredictable characteristics, are incorporated into hash functions to generate a consistent-length message digest from the original input. This study introduces a novel hash function that utilizes the logistic map with varying initial conditions and control parameters, making the most of the logistic map's chaotic range. By integrating the pseudo-random behavior within our compression function, predicting the output becomes exceedingly complex, inheriting the sensitive nature of the logistic map. The testing phase yielded highly favorable results when compared to recent research outcomes.
A Review of Clustering Algorithms for Big Data Kheyreddine Djouzi, Kadda Beghdad-Bey Proceedings Icnas 2019 4th International Conference on Networking and Advanced Systems, 2019 Big data is usually defined by five (05) characteristics called 5Vs +1C (Volume, Velocity, Variety, Veracity, Value and Complexity). It means to data that are too large, dynamic and complex with certain degree of accuracy. For that, data become difficult to analyze using traditional data analysis techniques because of their high complexity and computational cost. Clustering analysis technique is the most used method for cope with huge amount of data. The main goal of clustering is to classify data into clusters in manner that data grouped are more similar. In this paper, we provide an overview of various clustering techniques used for data analysis.
