We demonstrate that neural networks can be FLOP-efficient integrators of one-dimensional oscillatory integrands. We train a feed-forward neural network to compute integrals of highly oscillatory 1D functions. The training set is a parametric combination of functions with varying characters and oscillatory behavior degrees. Numerical examples show that these networks are FLOP-efficient for sufficiently oscillatory integrands with an average FLOP gain of 1000 FLOPs. The network calculates oscillatory integrals better than traditional quadrature methods under the same computational budget or number of floating point operations. We find that feed-forward networks of 5 hidden layers are satisfactory for a relative accuracy of 0.001. The computational burden of inference of the neural network is relatively small, even compared to inner-product pattern quadrature rules. We postulate that our result follows from learning latent patterns in the oscillatory integrands that are otherwise opaque to traditional numerical integrators.
相關內容
Networking：IFIP International Conferences on Networking。
Explanation：國(guo)際網(wang)絡會議。
Publisher：IFIP。
SIT：
The aim of the present work is to design, analyze theoretically, and test numerically, a generalized Dryja-Smith-Widlund (GDSW) preconditioner for composite Discontinuous Galerkin discretizations of multicompartment parabolic reaction-diffusion equations, where the solution can exhibit natural discontinuities across the domain. We prove that the resulting preconditioned operator for the solution of the discrete system arising at each time step converges with a scalable and quasi-optimal upper bound for the condition number. The GDSW preconditioner is then applied to the EMI (Extracellular - Membrane - Intracellular) reaction-diffusion system, recently proposed to model microscopically the spatiotemporal evolution of cardiac bioelectrical potentials. Numerical tests validate the scalability and quasi-optimality of the EMI-GDSW preconditioner, and investigate its robustness with respect to the time step size as well as jumps in the diffusion coefficients.
Video and wearable sensor data provide complementary information about human movement. Video provides a holistic understanding of the entire body in the world while wearable sensors provide high-resolution measurements of specific body segments. A robust method to fuse these modalities and obtain biomechanically accurate kinematics would have substantial utility for clinical assessment and monitoring. While multiple video-sensor fusion methods exist, most assume that a time-intensive, and often brittle, sensor-body calibration process has already been performed. In this work, we present a method to combine handheld smartphone video and uncalibrated wearable sensor data at their full temporal resolution. Our monocular, video-only, biomechanical reconstruction already performs well, with only several degrees of error at the knee during walking compared to markerless motion capture. Reconstructing from a fusion of video and wearable sensor data further reduces this error. We validate this in a mixture of people with no gait impairments, lower limb prosthesis users, and individuals with a history of stroke. We also show that sensor data allows tracking through periods of visual occlusion.
Accounting for uncertainty in Data quality is important for accurate statistical inference. We aim to an optimal conservative allocation for a large universe of assets in mean-variance portfolio (MVP), which is the worst choice within uncertainty in data distribution. Unlike the low dimensional MVP studied in Blanchet et al. (2022, Management Science), the large number of assets raises a challenging problem in quantifying the uncertainty, due to the big deviation of the sample covariance matrix from the population version. To overcome this difficulty, we propose a data-adaptive method to quantify the uncertainty with the help of a factor structure. Monte-Carlo Simulation is conducted to show the superiority of our method in high-dimensional cases, that, avoiding the over-conservative results in Blanchet et al. (2022), our allocation is closer to the oracle version in terms of risk minimization and expected portfolio return controlling.
Complex networks, which are the abstractions of many real-world systems, present a persistent challenge across disciplines for people to decipher their underlying information. Recently, hyperbolic geometry of latent spaces has gained traction in network analysis, due to its ability to preserve certain local intrinsic properties of the nodes. In this study, we explore the problem from a much broader perspective: understanding the impact of nodes' global topological structures on latent space placements. Our investigations reveal a direct correlation between the topological structure of nodes and their positioning within the latent space. Building on this deep and strong connection between node distance and network topology, we propose a novel embedding framework called Topology-encoded Latent Hyperbolic Geometry (TopoLa) for analyzing complex networks. With the encoded topological information in the latent space, TopoLa is capable of enhancing both conventional and low-rank networks, using the singular value gap to clarify the mathematical principles behind this enhancement. Meanwhile, we show that the equipped TopoLa distance can also help augment pivotal deep learning models encompassing knowledge distillation and contrastive learning.
With the recent popularity of neural networks comes the need for efficient serving of inference workloads. A neural network inference workload can be represented as a computational graph with nodes as operators transforming multidimensional tensors. The tensors can be transposed and/or tiled in a combinatorially large number of ways, some configurations leading to accelerated inference. We propose TGraph, a neural graph architecture that allows screening for fast configurations of the target computational graph, thus representing an artificial intelligence (AI) tensor compiler in contrast to the traditional heuristics-based compilers. The proposed solution improves mean Kendall's $\tau$ across layout collections of TpuGraphs from 29.8% of the reliable baseline to 67.4% of TGraph. We estimate the potential CO$_2$ emission reduction associated with our work to be equivalent to over 50% of the total household emissions in the areas hosting AI-oriented data centers.
Randomized controlled trials often suffer from interference, a violation of the Stable Unit Treatment Values Assumption (SUTVA) in which a unit's treatment assignment affects the outcomes of its neighbors. This interference causes bias in naive estimators of the average treatment effect (ATE). A popular method to achieve unbiasedness is to pair the Horvitz-Thompson estimator of the ATE with a known exposure mapping: a function that identifies which units in a given randomization are not subject to interference. For example, an exposure mapping can specify that any unit with at least $h$-fraction of its neighbors having the same treatment status does not experience interference. However, this threshold $h$ is difficult to elicit from domain experts, and a misspecified threshold can induce bias. In this work, we propose a data-adaptive method to select the "$h$"-fraction threshold that minimizes the mean squared error of the Hortvitz-Thompson estimator. Our method estimates the bias and variance of the Horvitz-Thompson estimator under different thresholds using a linear dose-response model of the potential outcomes. We present simulations illustrating that our method improves upon non-adaptive choices of the threshold. We further illustrate the performance of our estimator by running experiments on a publicly-available Amazon product similarity graph. Furthermore, we demonstrate that our method is robust to deviations from the linear potential outcomes model.
Sensorized insoles provide a tool for gait studies and health monitoring during daily life. For users to accept such insoles they need to be comfortable and lightweight. Previous work has already demonstrated that estimation of ground reaction forces (GRFs) is possible with insoles. However, these are often assemblies of commercial components restricting design freedom and customization. Within this work, we investigate using four 3D-printed soft foam-like sensors to sensorize an insole. These sensors were combined with system identification of Hammerstein-Wiener models to estimate the 3D GRFs, which were compared to values from an instrumented treadmill as the golden standard. It was observed that the four sensors behaved in line with the expected change in pressure distribution during the gait cycle. In addition, the identified (personalized) Hammerstein-Wiener models showed the best estimation performance (on average RMS error 9.3%, R^2=0.85 and mean absolute error (MAE) 7%) of the vertical, mediolateral, and anteroposterior GRFs. Thereby showing that these sensors can estimate the resulting 3D force reasonably well. These results for nine participants were comparable to or outperformed other works that used commercial FSRs with machine learning. The identified models did decrease in estimation performance over time but stayed on average 11.35% RMS and 8.6% MAE after a week with the Hammerstein-Wiener model seeming consistent between days two and seven. These results show promise for using 3D-printed soft piezoresistive foam-like sensors with system identification to be a viable approach for applications that require softness, lightweight, and customization such as wearable (force) sensors.
We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification. It is a simple residual network that alternates (i) a linear layer in which image patches interact, independently and identically across channels, and (ii) a two-layer feed-forward network in which channels interact independently per patch. When trained with a modern training strategy using heavy data-augmentation and optionally distillation, it attains surprisingly good accuracy/complexity trade-offs on ImageNet. We will share our code based on the Timm library and pre-trained models.
Graph neural networks (GNNs) have been proven to be effective in various network-related tasks. Most existing GNNs usually exploit the low-frequency signals of node features, which gives rise to one fundamental question: is the low-frequency information all we need in the real world applications? In this paper, we first present an experimental investigation assessing the roles of low-frequency and high-frequency signals, where the results clearly show that exploring low-frequency signal only is distant from learning an effective node representation in different scenarios. How can we adaptively learn more information beyond low-frequency information in GNNs? A well-informed answer can help GNNs enhance the adaptability. We tackle this challenge and propose a novel Frequency Adaptation Graph Convolutional Networks (FAGCN) with a self-gating mechanism, which can adaptively integrate different signals in the process of message passing. For a deeper understanding, we theoretically analyze the roles of low-frequency signals and high-frequency signals on learning node representations, which further explains why FAGCN can perform well on different types of networks. Extensive experiments on six real-world networks validate that FAGCN not only alleviates the over-smoothing problem, but also has advantages over the state-of-the-arts.
Recent advances in 3D fully convolutional networks (FCN) have made it feasible to produce dense voxel-wise predictions of volumetric images. In this work, we show that a multi-class 3D FCN trained on manually labeled CT scans of several anatomical structures (ranging from the large organs to thin vessels) can achieve competitive segmentation results, while avoiding the need for handcrafting features or training class-specific models. To this end, we propose a two-stage, coarse-to-fine approach that will first use a 3D FCN to roughly define a candidate region, which will then be used as input to a second 3D FCN. This reduces the number of voxels the second FCN has to classify to ~10% and allows it to focus on more detailed segmentation of the organs and vessels. We utilize training and validation sets consisting of 331 clinical CT images and test our models on a completely unseen data collection acquired at a different hospital that includes 150 CT scans, targeting three anatomical organs (liver, spleen, and pancreas). In challenging organs such as the pancreas, our cascaded approach improves the mean Dice score from 68.5 to 82.2%, achieving the highest reported average score on this dataset. We compare with a 2D FCN method on a separate dataset of 240 CT scans with 18 classes and achieve a significantly higher performance in small organs and vessels. Furthermore, we explore fine-tuning our models to different datasets. Our experiments illustrate the promise and robustness of current 3D FCN based semantic segmentation of medical images, achieving state-of-the-art results. Our code and trained models are available for download: //github.com/holgerroth/3Dunet_abdomen_cascade.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191