Results 91 to 100 of about 24,131 (257)

PS-SNN: pattern separation learning for expandable spiking neural networks in class-incremental learning

Scientific Reports
Biological brains mitigate interference by orthogonalizing neural representations of similar memories, thereby preserving stability across tasks in continual learning. However, most existing continual learning approaches for spiking neural networks (SNNs)
Ke Hu, Liangsheng Wen, Tingting Zhang, Hao Zhang   +3 more
doaj   +1 more source

Innovations in Gastric Cancer Surgery During Early Minimally Invasive Era and Future Perspectives

Annals of Gastroenterological Surgery, EarlyView.
With continuing revelations in tumor biology and the emergence of artificial intelligence, new horizons for surgical innovation are opening. At the center of this transformative journey stands the innovative surgeon, driven by passion, guided by data, and steadfast in the commitment to patient safety and quality of life.
Reut El‐On, Young‐Woo Kim
wiley   +1 more source

Challenges and enablers in fluidization technology

AIChE Journal, EarlyView.
Abstract Gas–solid fluidized beds provide excellent heat and mass transfer for high‐throughput operations from coating to catalytic conversion and underpin emerging low‐carbon technologies. Yet industrial reliability, scale‐up, and control lag scientific understanding, particularly as finer, stickier, and more variable feedstocks increasingly challenge
J. Ruud van Ommen, Jia Wei Chew
wiley   +1 more source

Feature-Generation-Replay Continual Learning Combined with Mixture-of-Experts for Data-Driven Autonomous Guidance

Drones
Continual learning (CL) is a key technology for enabling data-driven autonomous guidance systems to operate stably and persistently in complex and dynamic environments.
Bowen Li, Junxiang Li, Hongji Cheng, Tao Wu, Binhan Du   +4 more
doaj   +1 more source

Continual uncertainty learning

CoRR
Robust control of mechanical systems with multiple uncertainties remains a fundamental challenge, particularly when nonlinear dynamics and operating-condition variations are intricately intertwined. Although deep reinforcement learning (DRL) combined with domain randomization has shown promise in mitigating the sim-to-real gap, simultaneously handling ...
Heisei Yonezawa, Ansei Yonezawa, Itsuro Kajiwara   +2 more
openaire   +2 more sources

Eigenvector Continuation with Subspace Learning

Physical Review Letters, 2018
Version to appear in Physical Review Letters, 4 + 6 pages (main + supplemental materials), 1 + 6 figures (main + supplemental materials)
Frame, Dillon, He, Rongzheng, Ipsen, Ilse, Lee, Daniel, Lee, Dean, Rrapaj, Ermal   +5 more
openaire   +4 more sources

Factorization Machine‐Based Active Learning for Functional Materials Design with Optimal Initial Data

Advanced Intelligent Discovery, EarlyView.
This work investigates the optimal initial data size for surrogate‐based active learning in functional material optimization. Using factorization machine (FM)‐based quadratic unconstrained binary optimization (QUBO) surrogates and averaged piecewise linear regression, we show that adequate initial data accelerates convergence, enhances efficiency, and ...
Seongmin Kim, In‐Saeng Suh
wiley   +1 more source

When Biology Meets Medicine: A Perspective on Foundation Models

Advanced Intelligent Discovery, EarlyView.
Artificial intelligence, and foundation models in particular, are transforming life sciences and medicine. This perspective reviews biological and medical foundation models across scales, highlighting key challenges in data availability, model evaluation, and architectural design.
Kunying Niu, Yuxing Wang, Limin Zou, Jing Huang   +3 more
wiley   +1 more source

Hydration Behavior of Tricalcium Silicate in Seawater Relevant Salt Systems: A Hybrid Study with the Aid of Machine Learning

Advanced Intelligent Discovery, EarlyView.
The hydration behavior of C3S in seawater‐relevant solutions is studied based on experiments, boundary nucleation and growth (BNG) modeling, and machine learning. The main ions included in seawater modify hydration mechanisms, with MgCl2 showing the strongest acceleration effect at the same concentration.
Yanjie Sun, Chen Zhang, Ting Zhang, Mengze Hong, Peiliang Shen, Chi Sun Poon, Gang Li   +6 more
wiley   +1 more source

Open-world continual learning: Unifying novelty detection and continual learning

Artificial Intelligence
As AI agents are increasingly used in the real open world with unknowns or novelties, they need the ability to (1) recognize objects that (a) they have learned before and (b) detect items that they have never seen or learned, and (2) learn the new items incrementally to become more and more knowledgeable and powerful. (1) is called novelty detection or
Gyuhak Kim, Changnan Xiao, Tatsuya Konishi, Zixuan Ke, Bing Liu 0001   +4 more
openaire   +3 more sources