Machine unlearning using forgetting neural networks

conference paper Machine unlearning using forgetting neural networks published yes Amartya Hatua author Trung Nguyen author Filip Cano Cordoba author 708cad98-e86a-11ef-8098-bdae2d7c6af10000-0002-0783-904X Andrew Sung author ToHe department ICAART: International Conference on Agents and Artificial Intelligence Modern computer systems store vast amounts of personal data, enabling advances in AI and ML but risking user privacy and trust. For privacy reasons, it is sometimes desired for an ML model to forget part of the data it was trained on. In this paper, we introduce a novel unlearning approach based on Forgetting Neural Networks (FNNs), a neuroscience-inspired architecture that explicitly encodes forgetting through multiplicative decay factors. While FNNs had previously been studied as a theoretical construct, we provide the first concrete implementation and demonstrate their effectiveness for targeted unlearning. We propose several variants with per-neuron forgetting factors, including rank-based assignments guided by activation levels, and evaluate them on MNIST and Fashion-MNIST benchmarks. Our method systematically removes information associated with forget sets while preserving performance on retained data. Membership inference attacks confirm the effectiveness of FNN-based unlearning in erasing information about the training data from the neural network. These results establish FNNs as a promising foundation for efficient and interpretable unlearning. SciTePress2026Marbella, Spain eng Machine UnlearningNeuroscience-Inspired Machine LearningMembership Inference Attacks Proceedings of the 18th International Conference on Agents and Artificial Intelligence 2184-433X 9789897587962 2410.2237410.5220/0014326500004052 21536-1546 Hatua A, Nguyen T, Cano Cordoba F, Sung A. 2026. Machine unlearning using forgetting neural networks. Proceedings of the 18th International Conference on Agents and Artificial Intelligence. ICAART: International Conference on Agents and Artificial Intelligence vol. 2, 1536–1546. A. Hatua, T. Nguyen, F. Cano Cordoba, A. Sung, in:, Proceedings of the 18th International Conference on Agents and Artificial Intelligence, SciTePress, 2026, pp. 1536–1546. A. Hatua, T. Nguyen, F. Cano Cordoba, and A. Sung, “Machine unlearning using forgetting neural networks,” in Proceedings of the 18th International Conference on Agents and Artificial Intelligence, Marbella, Spain, 2026, vol. 2, pp. 1536–1546. Hatua, Amartya, Trung Nguyen, Filip Cano Cordoba, and Andrew Sung. “Machine Unlearning Using Forgetting Neural Networks.” In Proceedings of the 18th International Conference on Agents and Artificial Intelligence, 2:1536–46. SciTePress, 2026. <a href="https://doi.org/10.5220/0014326500004052">https://doi.org/10.5220/0014326500004052</a>. Hatua, A., Nguyen, T., Cano Cordoba, F., & Sung, A. (2026). Machine unlearning using forgetting neural networks. In Proceedings of the 18th International Conference on Agents and Artificial Intelligence (Vol. 2, pp. 1536–1546). Marbella, Spain: SciTePress. <a href="https://doi.org/10.5220/0014326500004052">https://doi.org/10.5220/0014326500004052</a> Hatua A, Nguyen T, Cano Cordoba F, Sung A. Machine unlearning using forgetting neural networks. In: Proceedings of the 18th International Conference on Agents and Artificial Intelligence. Vol 2. SciTePress; 2026:1536-1546. doi:<a href="https://doi.org/10.5220/0014326500004052">10.5220/0014326500004052</a> Hatua, Amartya, et al. “Machine Unlearning Using Forgetting Neural Networks.” Proceedings of the 18th International Conference on Agents and Artificial Intelligence, vol. 2, SciTePress, 2026, pp. 1536–46, doi:<a href="https://doi.org/10.5220/0014326500004052">10.5220/0014326500004052</a>. 222942026-07-13T09:46:46Z2026-07-16T09:02:53Z