Quantum genetic algorithm boosts anomaly detection with negative selection

Negative Selection Algorithms (NSAs) mimic the human immune system’s self/non-self discrimination to detect outliers, but their effectiveness often hinges on inefficient detector generation. In a new arXiv preprint (2605.22527), Giancarlo P. Gamberi and Calebe P. Bianchini propose the Quantum Genetic Negative Selection Algorithm (QGNSA), which integrates a Quantum Genetic Algorithm (QGA) into the existing EvoSeedRNSA framework. By exploiting quantum superposition and probabilistic amplitude adjustment, QGNSA dramatically enhances search space exploration and convergence speed compared to classical evolutionary methods. This approach addresses a core bottleneck in NSAs: generating a diverse and optimal set of detectors for anomaly detection.

Empirically, the team evaluated QGNSA on the Metaverse Financial Transactions Dataset – a high-dimensional, complex benchmark. The results show that QGNSA delivers superior detection accuracy while maintaining robustness across varying hyperparameter configurations. The authors attribute these gains to quantum parallelism that explores multiple candidate detector solutions simultaneously. This work highlights the potential of quantum computing in artificial immune systems, especially for cybersecurity and fraud detection. Future plans include optimizing quantum circuit designs, deploying on actual quantum hardware, and exploring hybrid quantum-classical architectures to further reduce computational overhead.