arXiv subjects

Ye He

Publications and source records attributed to Ye He.

30 records · Page 2Linked to original sources

arXiv2024

Regularized Stein Variational Gradient Flow

The Stein Variational Gradient Descent (SVGD) algorithm is a deterministic particle method for sampling. However, a mean-field analysis reveals that the gradient flow corresponding to the SVGD algorithm (i.e., the Stein Variational Gradient Flow) only provides a constant-order approximation to the Wasserstein Gradient Flow corresponding to the KL-divergence minimization. In this work, we propose the Regularized Stein Variational Gradient Flow, which interpolates between the Stein Variational Gradient Flow and the Wasserstein Gradient Flow. We establish various theoretical properties of the Regularized Stein Variational Gradient Flow (and its time-discretization) including convergence to equilibrium, existence and uniqueness of weak solutions, and stability of the solutions. We provide preliminary numerical evidence of the improved performance offered by the regularization.

stat.ML↗

arXiv2024

High-dimensional scaling limits and fluctuations of online least-squares SGD with smooth covariance

We derive high-dimensional scaling limits and fluctuations for the online least-squares Stochastic Gradient Descent (SGD) algorithm by taking the properties of the data generating model explicitly into consideration. Our approach treats the SGD iterates as an interacting particle system, where the expected interaction is characterized by the covariance structure of the input. Assuming smoothness conditions on moments of order up to eight orders, and without explicitly assuming Gaussianity, we establish the high-dimensional scaling limits and fluctuations in the form of infinite-dimensional Ordinary Differential Equations (ODEs) or Stochastic Differential Equations (SDEs). Our results reveal a precise three-step phase transition of the iterates; it goes from being ballistic, to diffusive, and finally to purely random behavior, as the noise variance goes from low, to moderate and finally to very-high noise setting. In the low-noise setting, we further characterize the precise fluctuations of the (scaled) iterates as infinite-dimensional SDEs. We also show the existence and uniqueness of solutions to the derived limiting ODEs and SDEs. Our results have several applications, including characterization of the limiting mean-square estimation or prediction errors and their fluctuations, which can be obtained by analytically or numerically solving the limiting equations.

math.PR↗

arXiv2024

High precision proton beam monitor system concept design on CSNS based on SiC

A high precision beam monitor system based on silicon carbide PIN sensor is designed for China Spallation Neutron Source 1.6 GeV proton beam to monitor the proton beam fluence.The concept design of the beam monitor system is finished together with front-end electronics with silicon carbide PIN sensors, readout system and mechanical system.Several tests are performed to study the performance of each component of the system.The charge collection of the SiC PIN sensors after proton radiation is studied with 80 MeV proton beam for continuous running. Research on the performance of the front-end electronics and readout system is finished for better data acquisition.The uncertainty of proton beam fluence is below 1% in the beam monitor system.

physics.acc-ph↗

arXiv2023

Towards a Complete Analysis of Langevin Monte Carlo: Beyond Poincaré Inequality

Langevin diffusions are rapidly convergent under appropriate functional inequality assumptions. Hence, it is natural to expect that with additional smoothness conditions to handle the discretization errors, their discretizations like the Langevin Monte Carlo (LMC) converge in a similar fashion. This research program was initiated by Vempala and Wibisono (2019), who established results under log-Sobolev inequalities. Chewi et al. (2022) extended the results to handle the case of Poincaré inequalities. In this paper, we go beyond Poincaré inequalities, and push this research program to its limit. We do so by establishing upper and lower bounds for Langevin diffusions and LMC under weak Poincaré inequalities that are satisfied by a large class of densities including polynomially-decaying heavy-tailed densities (i.e., Cauchy-type). Our results explicitly quantify the effect of the initializer on the performance of the LMC algorithm. In particular, we show that as the tail goes from sub-Gaussian, to sub-exponential, and finally to Cauchy-like, the dependency on the initial error goes from being logarithmic, to polynomial, and then finally to being exponential. This three-step phase transition is in particular unavoidable as demonstrated by our lower bounds, clearly defining the boundaries of LMC.

math.ST↗

arXiv2023

Mean-Square Analysis of Discretized Itô Diffusions for Heavy-tailed Sampling

We analyze the complexity of sampling from a class of heavy-tailed distributions by discretizing a natural class of Itô diffusions associated with weighted Poincaré inequalities. Based on a mean-square analysis, we establish the iteration complexity for obtaining a sample whose distribution is $ε$ close to the target distribution in the Wasserstein-2 metric. In this paper, our results take the mean-square analysis to its limits, i.e., we invariably only require that the target density has finite variance, the minimal requirement for a mean-square analysis. To obtain explicit estimates, we compute upper bounds on certain moments associated with heavy-tailed targets under various assumptions. We also provide similar iteration complexity results for the case where only function evaluations of the unnormalized target density are available by estimating the gradients using a Gaussian smoothing technique. We provide illustrative examples based on the multivariate $t$-distribution.

math.ST↗

arXiv2022

Heavy-tailed Sampling via Transformed Unadjusted Langevin Algorithm

We analyze the oracle complexity of sampling from polynomially decaying heavy-tailed target densities based on running the Unadjusted Langevin Algorithm on certain transformed versions of the target density. The specific class of closed-form transformation maps that we construct are shown to be diffeomorphisms, and are particularly suited for developing efficient diffusion-based samplers. We characterize the precise class of heavy-tailed densities for which polynomial-order oracle complexities (in dimension and inverse target accuracy) could be obtained, and provide illustrative examples. We highlight the relationship between our assumptions and functional inequalities (super and weak Poincaré inequalities) based on non-local Dirichlet forms defined via fractional Laplacian operators, used to characterize the heavy-tailed equilibrium densities of certain stable-driven stochastic differential equations.

math.ST↗

arXiv2021

On the Ergodicity, Bias and Asymptotic Normality of Randomized Midpoint Sampling Method

The randomized midpoint method, proposed by [SL19], has emerged as an optimal discretization procedure for simulating the continuous time Langevin diffusions. Focusing on the case of strong-convex and smooth potentials, in this paper, we analyze several probabilistic properties of the randomized midpoint discretization method for both overdamped and underdamped Langevin diffusions. We first characterize the stationary distribution of the discrete chain obtained with constant step-size discretization and show that it is biased away from the target distribution. Notably, the step-size needs to go to zero to obtain asymptotic unbiasedness. Next, we establish the asymptotic normality for numerical integration using the randomized midpoint method and highlight the relative advantages and disadvantages over other discretizations. Our results collectively provide several insights into the behavior of the randomized midpoint discretization method, including obtaining confidence intervals for numerical integrations.

stat.ML↗

arXiv2020

Mutual-Supervised Feature Modulation Network for Occluded Pedestrian Detection

State-of-the-art pedestrian detectors have achieved significant progress on non-occluded pedestrians, yet they are still struggling under heavy occlusions. The recent occlusion handling strategy of popular two-stage approaches is to build a two-branch architecture with the help of additional visible body annotations. Nonetheless, these methods still have some weaknesses. Either the two branches are trained independently with only score-level fusion, which cannot guarantee the detectors to learn robust enough pedestrian features. Or the attention mechanisms are exploited to only emphasize on the visible body features. However, the visible body features of heavily occluded pedestrians are concentrated on a relatively small area, which will easily cause missing detections. To address the above issues, we propose in this paper a novel Mutual-Supervised Feature Modulation (MSFM) network, to better handle occluded pedestrian detection. The key MSFM module in our network calculates the similarity loss of full body boxes and visible body boxes corresponding to the same pedestrian so that the full-body detector could learn more complete and robust pedestrian features with the assist of contextual features from the occluding parts. To facilitate the MSFM module, we also propose a novel two-branch architecture, consisting of a standard full body detection branch and an extra visible body classification branch. These two branches are trained in a mutual-supervised way with full body annotations and visible body annotations, respectively. To verify the effectiveness of our proposed method, extensive experiments are conducted on two challenging pedestrian datasets: Caltech and CityPersons, and our approach achieves superior performance compared to other state-of-the-art methods on both datasets, especially in heavy occlusion case.

cs.CV↗

arXiv2017

Fluctuations between high- and low-modularity topology in time-resolved functional connectivity

Modularity is an important topological attribute for functional brain networks. Recent studies have reported that modularity of functional networks varies not only across individuals being related to demographics and cognitive performance, but also within individuals co-occurring with fluctuations in network properties of functional connectivity, estimated over short time intervals. However, characteristics of these time-resolved functional networks during periods of high and low modularity have remained largely unexplored. In this study we investigate spatiotemporal properties of time-resolved networks in the high and low modularity periods during rest, with a particular focus on their spatial connectivity patterns, temporal homogeneity and test-retest reliability. We show that spatial connectivity patterns of time-resolved networks in the high and low modularity periods are represented by increased and decreased dissociation of the default mode network module from task-positive network modules, respectively. We also find that the instances of time-resolved functional connectivity sampled from within the high (low) modularity period are relatively homogeneous (heterogeneous) over time, indicating that during the low modularity period the default mode network interacts with other networks in a variable manner. We confirmed that the occurrence of the high and low modularity periods varies across individuals with moderate inter-session test-retest reliability and that it is correlated with previously-reported individual differences in the modularity of functional connectivity estimated over longer timescales. Our findings illustrate how time-resolved functional networks are spatiotemporally organized during periods of high and low modularity, allowing one to trace individual differences in long-timescale modularity to the variable occurrence of network configurations at shorter timescales.

q-bio.NC↗

arXiv2015

Dynamic fluctuations coincide with periods of high and low modularity in resting-state functional brain networks

We investigate the relationship of resting-state fMRI functional connectivity estimated over long periods of time with time-varying functional connectivity estimated over shorter time intervals. We show that using Pearson's correlation to estimate functional connectivity implies that the range of fluctuations of functional connections over short time scales is subject to statistical constraints imposed by their connectivity strength over longer scales. We present a method for estimating time-varying functional connectivity that is designed to mitigate this issue and allows us to identify episodes where functional connections are unexpectedly strong or weak. We apply this method to data recorded from $N=80$ participants, and show that the number of unexpectedly strong/weak connections fluctuates over time, and that these variations coincide with intermittent periods of high and low modularity in time-varying functional connectivity. We also find that during periods of relative quiescence regions associated with default mode network tend to join communities with attentional, control, and primary sensory systems. In contrast, during periods where many connections are unexpectedly strong/weak, default mode regions dissociate and form distinct modules. Finally, we go on to show that, while all functional connections can at times manifest stronger (more positively correlated) or weaker (more negatively correlated) than expected, a small number of connections, mostly within the visual and somatomotor networks, do so a disproportional number of times. Our statistical approach allows the detection of functional connections that fluctuate more or less than expected based on their long-time averages and may be of use in future studies characterizing the spatio-temporal patterns of time-varying functional connectivity

q-bio.NC↗

arXiv2015

Functional brain modules reconfigure at multiple scales across the human lifespan

The human brain is a complex network of interconnected brain regions organized into functional modules with distinct roles in cognition and behavior. An important question concerns the persistence and stability of these modules over the human lifespan. Here we use graph-theoretic analysis to algorithmically uncover the brain's intrinsic modular organization across multiple spatial scales ranging from small communities comprised of only a few brain regions to large communities made up of many regions. We find that at coarse scales modules become progressively more segregated, while at finer scales segregation decreases. Module composition also exhibits scale-specific and age-dependent changes. At coarse scales, the module assignments of regions normally associated with control, default mode, attention, and visual networks are highly flexible. At fine scales the most flexible regions are associated with the default mode network. Finally, we show that, with age, some regions in the default mode network, specifically retrosplenial cortex, maintain a greater proportion of functional connections to their own module, while regions associated with somatomotor and saliency/ventral attention networks distribute their links more evenly across modules.

q-bio.NC↗

arXiv2015

Generative models of the human connectome

The human connectome represents a network map of the brain's wiring diagram and the pattern into which its connections are organized is thought to play an important role in cognitive function. The generative rules that shape the topology of the human connectome remain incompletely understood. Earlier work in model organisms has suggested that wiring rules based on geometric relationships (distance) can account for many but likely not all topological features. Here we systematically explore a family of generative models of the human connectome that yield synthetic networks designed according to different wiring rules combining geometric and a broad range of topological factors. We find that a combination of geometric constraints with a homophilic attachment mechanism can create synthetic networks that closely match many topological characteristics of individual human connectomes, including features that were not included in the optimization of the generative model itself. We use these models to investigate a lifespan dataset and show that, with age, the model parameters undergo progressive changes, suggesting a rebalancing of the generative factors underlying the connectome across the lifespan.

q-bio.NC↗