Results 131 to 140 of about 6,336,923 (296)

Continual graph learning: A survey

Pattern Recognition
Continual Graph Learning (CGL) enables models to incrementally learn from streaming graph-structured data without forgetting previously acquired knowledge. Experience replay is a common solution that reuses a subset of past samples during training. However, it may lead to information loss and privacy risks. Generative replay addresses these concerns by
Qiao Yuan, Sheng-Uei Guan, Pin Ni, Tianlun Luo, Prudence Wong, Victor Chang, Ka Lok Man   +6 more
openaire   +4 more sources

continual learning strategies for anomaly detection

openThis thesis investigates the application of continual learning to anomaly detection, aiming to develop models that can adapt to new data while preserving previously acquired knowledge. A major challenge is catastrophic forgetting — when a model loses
RAHIMI, MAHAN
core  

Continual learning from a probabilistic perspective

Over the past few years there has been great advancements in machine learning. It is now possible to learn models that predict some quantity of interest with high accuracy.
Lee, Thomas L.
core   +1 more source

The Collective Power of Bacteria as a Blueprint for Emergent Intelligence

Advanced Intelligent Systems, EarlyView.
Small cells, powerful collectives. Bacteria demonstrate how sophisticated behaviors can emerge from many simple individuals working together. We explore the remarkable world of bacterial communities and the mechanisms that underpin their complex emergent behaviors, enabling impressive adaptability, robustness, and responsiveness to changing ...
Johanna A. Blee, Thomas E. Gorochowski, Sabine Hauert   +2 more
wiley   +1 more source

Outlook towards deployable continual learning for particle accelerators

Machine Learning: Science and Technology
Particle accelerators are high power complex machines. To ensure uninterrupted operation of these machines, thousands of pieces of equipment need to be synchronized, which requires addressing many challenges including design, optimization and control ...
Kishansingh Rajput, Sen Lin, Auralee Edelen, Willem Blokland, Malachi Schram   +4 more
doaj   +1 more source

Adapt Before Continual Learning

Proceedings of the AAAI Conference on Artificial Intelligence
Continual Learning (CL) seeks to enable neural networks to incrementally acquire new knowledge (plasticity) while retaining existing knowledge (stability). Although pre-trained models (PTMs) have provided a strong foundation for CL, existing approaches face a fundamental challenge in balancing these two competing objectives.
Aojun Lu, Tao Feng 0014, Hangjie Yuan, Chunhui Ding, Yanan Sun 0001   +4 more
openaire   +4 more sources

Resilient Multimodal Fusion for Robust Vulnerability Exploitability Prediction Under Data Degradation

Advanced Intelligent Systems, EarlyView.
Secure Fusion‐X harmonizes unstructured NVD descriptions with structured CVSS/CWE/CPE metadata via decision‐level fusion, overcoming the fragility of traditional unimodal models. Automated assessment of software vulnerability exploitability is essential for intelligent cyber defense, yet its effectiveness is often hindered by unstable, delayed, or ...
Mona Dolati, Seyedeh Leili Mirtaheri, Reza Shahbazian, Andrea Pugliese   +3 more
wiley   +1 more source

Continual learning in the presence of repetition

Neural Networks
Continual learning (CL) provides a framework for training models in ever-evolving environments. Although re-occurrence of previously seen objects or tasks is common in real-world problems, the concept of repetition in the data stream is not often considered in standard benchmarks for CL.
Hamed Hemati, Lorenzo Pellegrini, Xiaotian Duan, Zixuan Zhao, Fangfang Xia, Marc Masana, Benedikt Tscheschner, Eduardo E. Veas, Yuxiang Zheng, Shiji Zhao, Shao-Yuan Li, Sheng-Jun Huang, Vincenzo Lomonaco, Gido M. van de Ven   +13 more
openaire   +5 more sources

From Instruction to Inheritance: Scaling Robot Learning Through Knowledge Circulation

Advanced Intelligent Systems, EarlyView.
Robot learning moves from one‐way instruction toward knowledge circulation. Humans teach robots, robots teach one another, and robots share acquired skills with humans. This perspective connects teaching, embodiment, and learning architectures to explore how knowledge is inherited, adapted, and enriched across agents, opening a path toward scalable ...
Kento Kawaharazuka, Shunki Itadera, Kohei Honda, Takato Horii, Asako Kanezaki, Taisuke Kobayashi, Kenji Koide, Shuhei Kurita, Koshi Makihara, Kazuki Miyazawa, Masaki Murooka, Yuki Onishi, Satoshi Yagi, Satoshi Yamamori   +13 more
wiley   +1 more source

Freedom dreaming of migrant justice: Critical reflections on counterspaces and institutional violence in the university

American Journal of Community Psychology, EarlyView.
Abstract US universities are built on stolen land and sustained through hierarchies of power that produce what migrant justice scholars name as b/order regimes. As institutions that claim to be sites of learning and inclusion, universities are fraught with contradictions as simultaneously sites of dispossession, exclusion, and control.
Sara L. Buckingham, Jesica Siham Fernández   +1 more
wiley   +1 more source