analiza aiqbench
CIFNet introduces a deterministic, closed-form approach to class-incremental learning that eliminates the need for iterative backpropagation and gradient updates. By leveraging frozen pre-trained representations with regularized recursive least-squares, this framework significantly reduces computational costs while preventing catastrophic forgetting across standard benchmarks like CIFAR-100 and ImageNet-100. The method achieves up to a 20-fold reduction in energy consumption by replacing epoch-wise optimization with exact analytical solutions for classifier weight adaptation.
CIFNet: An Analytic Neural Learning Framework for Efficient and Calibrated Class-Incremental Learning
arXiv:2509.11285v2 Announce Type: replace-cross Abstract: Class-Incremental Learning (CIL) in deep neural networks is conventionally framed as an iterative gradient-based optimization problem, incurring high computational cost, hyperparameter sensitivity, and risk of catastrophic forgetting. In this work, we demonstrate that when leveraging frozen pre-trained representations, CIL can be solved as a sequence of deterministic, closed-form classifier adaptations without backpropagation or iterative convergence. We propose CIFNet, an analytic neural learning framework built upon Regularised Recursive Least-Squares (RRLS). CIFNet updates classifier weights via an exact, closed-form ridge-regression solution operating in a stationary embedding space. To counteract the structural initialisation bias that arises when newly expanded output neurons are introduced without exposure to past-class evidence, CIFNet incorporates a lightweight calibration buffer in latent space alongside density-aware oversampling, ensuring globally balanced decision boundaries without raw image storage or gradient updates. Extensive evaluations across CIFAR-100, ImageNet-100, and CORe50 show that CIFNet achieves predictive accuracy competitive with iterative CIL baselines while maintaining strictly monotonic, smooth learning trajectories free from intermediate performance collapse. Furthermore, by replacing epoch-wise backpropagation with closed-form moment accumulation, CIFNet achieves up to 20$\times$ reduction in energy consumption. These findings establish calibrated analytic learning as an efficient, stable, and mathematically grounded paradigm for continual adaptation in neural networks.
Abstract:Class-Incremental Learning (CIL) in deep neural networks is conventionally framed as an iterative gradient-based optimization problem, incurring high computational cost, hyperparameter sensitivity, and risk of catastrophic forgetting. In this work, we demonstrate that when leveraging frozen pre-trained representations, CIL can be solved as a sequence of deterministic, closed-form classifier adaptations without backpropagation or iterative convergence. We propose CIFNet, an analytic neural learning framework built upon Regularised Recursive Least-Squares (RRLS). CIFNet updates classifier weights via an exact, closed-form ridge-regression solution operating in a stationary embedding space. To counteract the structural initialisation bias that arises when newly expanded output neurons are introduced without exposure to past-class evidence, CIFNet incorporates a lightweight calibration buffer in latent space alongside density-aware oversampling, ensuring globally balanced decision boundaries without raw image storage or gradient updates. Extensive evaluations across CIFAR-100, ImageNet-100, and CORe50 show that CIFNet achieves predictive accuracy competitive with iterative CIL baselines while maintaining strictly monotonic, smooth learning trajectories free from intermediate performance collapse. Furthermore, by replacing epoch-wise backpropagation with closed-form moment accumulation, CIFNet achieves up to 20$\times$ reduction in energy consumption. These findings establish calibrated analytic learning as an efficient, stable, and mathematically grounded paradigm for continual adaptation in neural networks.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2509.11285 [cs.LG]
(or arXiv:2509.11285v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2509.11285 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 292 cuvinte
Subiecte conexe
News
Quantum Computing Report · 25 iul. 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · 17 sept. 2026
Handling Imbalanced Classification: What Works Better Than SMOTE
Analytics Vidhya · 12 iul. 2026
[AINews] not much happened today
Latent Space · 3 oct. 2026