In many applications in engineering and sciences analysts have simultaneous access to multiple data sources. In such cases, the overall cost of acquiring information can be reduced via data fusion or multi-fidelity (MF) modeling where one leverages inexpensive low-fidelity (LF) sources to reduce the reliance on expensive high-fidelity (HF) data. In this paper, we employ neural networks (NNs) for data fusion in scenarios where data is very scarce and obtained from an arbitrary number of sources with varying levels of fidelity and cost. We introduce a unique NN architecture that converts MF modeling into a nonlinear manifold learning problem. Our NN architecture inversely learns non-trivial (e.g., non-additive and non-hierarchical) biases of the LF sources in an interpretable and visualizable manifold where each data source is encoded via a low-dimensional distribution. This probabilistic manifold quantifies model form uncertainties such that LF sources with small bias are encoded close to the HF source. Additionally, we endow the output of our NN with a parametric distribution not only to quantify aleatoric uncertainties, but also to reformulate the network's loss function based on strictly proper scoring rules which improve robustness and accuracy on unseen HF data. Through a set of analytic and engineering examples, we demonstrate that our approach provides a high predictive power while quantifying various sources uncertainties.
Credibility-Aware Multi-Modal Fusion Using Probabilistic Circuits
Erik Jonsson School of Engineering & Computer Science, The University of Texas at Dallas（德克萨斯大学达拉斯分校埃里克·琼森工程与计算机科学学院）；Air Force Research Laboratory（空军研究实验室）；TU Darmstadt（达姆施塔特工业大学）
Multi-fidelity aerodynamic data fusion by autoencoder transfer learning
Department of Aerospace Engineering, UC3M（航空航天工程系，UC3M）；Theoretical and Computational Aerodynamics Group, Flight Physics Department, INTA（理论与计算空气动力学组，飞行物理部门，INTA）
Fully probabilistic design for knowledge fusion between Bayesian filters under uniform disturbances
Czech Academy of Sciences, Institute of Information Theory and Automation（捷克科学院信息理论与自动化研究所）；Trinity College Dublin, the University of Dublin（都柏林大学三一学院）
Fidelity-based Probabilistic Q-learning for Control of Quantum Systems
Nanjing University（南京大学）；Princeton University（普林斯顿大学）；University of New South Wales（新南威尔士大学）；Zhejiang University（浙江大学）；City University of Hong Kong（香港城市大学）；Washington University in St. Louis（圣路易斯华盛顿大学）
LPCVAE: A Conditional VAE with Long-Term Dependency and Probabilistic Time-Frequency Fusion for Time Series Anomaly Detection
Institute of Information Engineering, Chinese Academy of Sciences, Beijing, China（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences, Beijing, China（中国科学院大学网络安全学院）
GAN-MDF: A Method for Multi-fidelity Data Fusion in Digital Twins
Dalian University of Technology（大连理工大学）；JD Explore Academy（京东探索研究院）
物联网（IoT）收集物理系统的实时数据，如智能工厂、智能机器人和医疗保健系统，并为数字孪生提供必要的支持。根据质量和精度，这些多源数据被划分为不同的保真度级别。高保真（HF）响应准确地描述了感兴趣的系统，但计算成本高。相比之下，低保真度（LF）响应具有较低的计算成本，但不能满足所需的精度。多保真度数据融合（Multi-fidelity data fusion，简称MIMO）方法的目标是利用大量低频样本和少量高频样本，以合理的计算量建立一个准确、高效的模型来描述系统。在本文中，我们提出了一种新的生成对抗网络的数字孪生（GAN-网络）。GAN-SVM的生成器由两个子网络组成：一个从输入中提取LF特征;另一个将输入和提取的LF特征集成以形成后续SVM的输入。GAN-SVM的样本识别生成器输出是否是从HF模型生成的真实样本。为了增强GAN-SVM训练的稳定性，我们还引入了监督损失技巧，以在对抗训练的每次迭代中细化生成器权重。与现有方法相比，该方法具有以下优点：1）在嵌套或非嵌套样本结构下均表现良好; 2）对数据分布没有特定的假设; 3）即使在HF样本很少的情况下也具有很高的鲁棒性。实验结果也支持了GAN-SVM的有效性。
Personalized Treatment Outcome Prediction from Scarce Data via Dual-Channel Knowledge Distillation and Adaptive Fusion
School of Information Management, Central China Normal University（中央财经大学信息管理学院）；School of Cyber Science and Engineering, Southeast University（东南大学计算机科学与工程学院）；Guangdong Provincial Key Laboratory of Brain-inspired Intelligent Computation, Department of Computer Science and Engineering, Southern University of Science and Technology（广东省脑启发式智能计算重点实验室，南方科技大学计算机科学与工程学院）
Probabilistic Token Alignment for Large Language Model Fusion
Rochester Institute of Technology（罗切斯特理工学院）；U.S. Naval Research Laboratory（美国海军研究实验室）；University of Missouri-Kansas City（密苏里大学-Kansas城分校）；Adobe（Adobe公司）；Meituan（美团）；Sun Yat-sen University（孙中山大学）；Purdue University（普渡大学）；UC Davis（加州大学戴维斯分校）；University of Rochester（罗切斯特大学）；Rice University（莱斯大学）；Meta AI
Multifidelity data fusion in convolutional encoder/decoder networks
University of Notre Dame（圣母大学）；Sandia National Laboratories（桑迪亚国家实验室）；Stanford University（斯坦福大学）
我们分析了由编码器、解码器和跳过连接组装并使用多保真度数据训练的卷积神经网络的回归精度。除了需要比等效的全连接网络少得多的可训练参数之外，编码器、解码器、编码器-解码器或解码器-编码器架构可以学习任意维度的输入到输出之间的映射。我们证明了他们的准确性时，训练的一些高保真度和许多低保真度的数据从模型，从一维函数到泊松方程求解器在二维。最后，我们讨论了一些实施方案的选择，提高了Monte Carlo DropBlocks产生的不确定性估计的可靠性，并比较低，高和多保真度的方法之间的不确定性估计。
Multi-Path Long-Term Vessel Trajectories Forecasting with Probabilistic Feature Fusion for Problem Shifting
organization= Institute for Big Data Analytics, Dalhousie University , addressline= Halifax -- NS, Canada；organization= Industrial Engineering Department, Dalhousie University, Halifax -- NS, Canada；organization= Department of Comp. Sci.；Media Tech., Linnaeus University , addressline= Vaxjo, Sweden；organization= Institute of Computer Science, Polish Academy of Sciences , addressline= Warsaw, Poland；organization= DHI Water \& Environment Inc. , addressline= Ottawa, Canada
NavMapFusion: Diffusion-based Fusion of Navigation Maps for Online Vectorized HD Map Construction
Mercedes-Benz Research & Development North America, USA（梅赛德斯-奔驰北美研究与开发）；University of Stuttgart, Germany（斯图加特大学）；University of California, San Diego, USA（加州大学圣地亚哥分校）；University of Southampton, United Kingdom（南安普顿大学）
