The rapid growth of research in exploiting machine learning to predict chaotic systems has revived a recent interest in Hamiltonian Neural Networks (HNNs) with physical constraints defined by the Hamilton's equations of motion, which represent a major class of physics-enhanced neural networks. We introduce a class of HNNs capable of adaptable prediction of nonlinear physical systems: by training the neural network based on time series from a small number of bifurcation-parameter values of the target Hamiltonian system, the HNN can predict the dynamical states at other parameter values, where the network has not been exposed to any information about the system at these parameter values. The architecture of the HNN differs from the previous ones in that we incorporate an input parameter channel, rendering the HNN parameter--cognizant. We demonstrate, using paradigmatic Hamiltonian systems, that training the HNN using time series from as few as four parameter values bestows the neural machine with the ability to predict the state of the target system in an entire parameter interval. Utilizing the ensemble maximum Lyapunov exponent and the alignment index as indicators, we show that our parameter-cognizant HNN can successfully predict the route of transition to chaos. Physics-enhanced machine learning is a forefront area of research, and our adaptable HNNs provide an approach to understanding machine learning with broad applications.
An Interpretable Neural Control Network with Adaptable Online Learning for Sample Efficient Robot Locomotion Learning
Vidyasirimedhi Institute of Science and Technology（维达亚希米德科学与技术学院）；The Mærsk Mc-Kinney Møller Institute（马士基麦金尼·莫勒研究所）；University of Southern Denmark（南方丹麦大学）
Test-Time Adaptable Neural Networks for Robust Medical Image Segmentation
ETH Zurich（苏黎世联邦理工学院）
当训练数据集代表测试时预期会遇到的变化时，卷积神经网络（CNN）对于监督学习问题非常有效。在医学图像分割中，当训练图像和测试图像在其采集细节（例如扫描仪模型或协议）方面不匹配时，这一前提就被违反了。在这种情况下，CNN的显着性能下降在文献中有很好的记录。为了解决这个问题，我们将分割CNN设计为两个子网络的级联：一个相对较浅的图像归一化CNN，然后是一个对归一化图像进行分割的深度CNN。我们使用训练数据集训练这两个子网络，该训练数据集由来自特定扫描仪和协议设置的注释图像组成。现在，在测试时，我们在预测分割标签的隐式先验的指导下，调整\emph{each test image}的图像归一化子网络。我们采用了一个独立训练的去噪自动编码器（DAE），以模拟这样一个隐含的前合理的解剖分割标签。我们在三个解剖结构的多中心磁共振成像数据集上验证了所提出的想法：大脑，心脏和前列腺。所提出的测试时间适应一致地提供性能改进，展示了该方法的承诺和通用性。由于不知道第二子网络深度CNN的架构，因此所提出的设计可以与任何分割网络一起使用，以提高对成像扫描仪和协议变化的鲁棒性。我们的代码可在：\url{https://github.com/neerakara/test-time-adaptable-neural-networks-for-domain-generalization} .
Adaptable Hindsight Experience Replay for Search-Based Learning
Technical University of Darmstadt（德累斯顿技术大学）；Hessian Center for Artificial Intelligence（黑森人工智能中心）；German Research Center for Artificial Intelligence（德国人工智能研究中心）；Centre for Cognitive Science（认知科学中心）；National Research Center for Applied Cybersecurity ATHENE（应用网络安全国家研究中心ATHENE）
本文提出 Adaptable HER 框架，将事后经验回放与 AlphaZero 式搜索学习灵活集成，通过可调整的重新标记目标、策略目标和轨迹选择，在稀疏奖励及方程发现等任务中超越纯监督学习和强化学习。
Machine Learning Hamiltonian Dynamical Systems with Sparse and Noisy Data
The Mathematical Institute, University of Oxford, Oxford OX2 6GG, United Kingdom（牛津大学数学研究所）；Center for Theoretical Physics, St. Stephen's College, University of Delhi, Delhi 110007, India（德里大学斯特金学院理论物理中心）
SA-GAT-SR: Self-Adaptable Graph Attention Networks with Symbolic Regression for high-fidelity material property prediction
University of Electronic Science and Technology of China（电子科技大学）；Tsinghua University（清华大学）；Los Alamos National Laboratory（洛斯阿拉莫斯国家实验室）；State Key Laboratory of Low Dimensional Quantum Physics（低维量子物理国家重点实验室）；Institute for Advanced Study（高级研究院）；Frontier Science Center for Quantum Information（量子信息前沿科学中心）
A General Framework for Interpretable Neural Learning based on Local Information-Theoretic Goal Functions
University of Göttingen（哥廷根大学）；Max Planck Institute for Dynamics and Self-Organization（马克斯·普朗克动力学与自组织研究所）；ETH Zurich（苏黎世联邦理工学院）；Champalimaud Centre for the Unknown（尚帕利莫未知中心）
Symplectic Neural Networks for Learning Non-Separable Hamiltonians
Czech Technical University in Prague（捷克布拉格理工大学）；University of California at San Diego（加州大学圣地亚哥分校）；Athena Research and Innovation Center（雅典研究与创新中心）；HSBC Holdings（汇丰控股）
Learning partially observed systems with neural Hamiltonian ordinary differential equations
Department of Mathematics and Cybernetics, SINTEF Digital（数学与自动化系，SINTEF数字研究院）；Department of Physics, University of Oslo（物理系，奥斯陆大学）；Department of Data Science, Norwegian University of Life Science（数据科学系，挪威生命科学大学）
Port-Hamiltonian Neural Networks with Output Error Noise Models
Eindhoven University of Technology（埃因霍温理工大学）；Eindhoven Artificial Intelligence Systems Institute（埃因霍温人工智能系统研究所）；HUN-REN Institute for Computer Science and Control（匈牙利科学院计算机科学与控制研究所（HUN-REN））
A Hamiltonian driven Geometric Construction of Neural Networks via the Lognormal family, Application to Financial Fraud Detection and to Network Security
University of Yaounde1（雅温迪大学）；University of Maroua（马拉乌大学）；Higher national school of Polytechnic（国家高等理工学院）
Learning Stochastic Hamiltonian Systems via Stochastic Generating Function Neural Network
School of Mathematical Sciences, University of Chinese Academy of Sciences（中国科学院大学数学科学学院）；Department of Mathematics and Statistics, Auburn University（阿伯茨罕大学数学与统计学系）
