no code implementations • NeurIPS 2014 • Xu Chen, Xiuyuan Cheng, Stéphane Mallat
The classification of high-dimensional data defined on graphs is particularly difficult when the graph geometry is unknown.
no code implementations • 30 Sep 2015 • Xiuyuan Cheng, Xu Chen, Stephane Mallat
An orthogonal Haar scattering transform is a deep network, computed with a hierarchy of additions, subtractions and absolute values, over pairs of coefficients.
1 code implementation • 6 Feb 2016 • Uri Shaham, Xiuyuan Cheng, Omer Dror, Ariel Jaffe, Boaz Nadler, Joseph Chang, Yuval Kluger
We show how deep learning methods can be applied in the context of crowdsourcing and unsupervised ensemble learning.
no code implementations • 9 Nov 2016 • Xiuyuan Cheng, Manas Rachh, Stefan Steinerberger
We study directed, weighted graphs $G=(V, E)$ and consider the (not necessarily symmetric) averaging operator $$ (\mathcal{L}u)(i) = -\sum_{j \sim_{} i}{p_{ij} (u(j) - u(i))},$$ where $p_{ij}$ are normalized edge weights.
no code implementations • 24 May 2017 • Bowei Yan, Purnamrita Sarkar, Xiuyuan Cheng
Community detection is a fundamental unsupervised learning problem for unlabeled networks which has a broad range of applications.
no code implementations • 5 Jun 2017 • Xiuyuan Cheng, Gal Mishne, Stefan Steinerberger
Let $(M, g)$ be a compact manifold and let $-\Delta \phi_k = \lambda_k \phi_k$ be the sequence of Laplacian eigenfunctions.
1 code implementation • 14 Sep 2017 • Xiuyuan Cheng, Alexander Cloninger, Ronald R. Coifman
The paper introduces a new kernel-based Maximum Mean Discrepancy (MMD) statistic for measuring the distance between two distributions given finitely-many multivariate samples.
1 code implementation • ICML 2018 • Qiang Qiu, Xiuyuan Cheng, Robert Calderbank, Guillermo Sapiro
In this paper, we suggest to decompose convolutional filters in CNN as a truncated expansion with pre-fixed bases, namely the Decomposed Convolutional Filters network (DCFNet), where the expansion coefficients remain learned from data.
1 code implementation • 28 Mar 2018 • Uri Shaham, James Garritano, Yutaro Yamada, Ethan Weinberger, Alex Cloninger, Xiuyuan Cheng, Kelly Stanton, Yuval Kluger
We study the effectiveness of various approaches that defend against adversarial attacks on deep networks via manipulations based on basis function representations of images.
no code implementations • ICLR 2019 • Xiuyuan Cheng, Qiang Qiu, Robert Calderbank, Guillermo Sapiro
Explicit encoding of group actions in deep features makes it possible for convolutional neural networks (CNNs) to handle global deformations of images, which is critical to success in many vision tasks.
1 code implementation • 18 May 2018 • Yingzhou Li, Xiuyuan Cheng, Jianfeng Lu
Theoretical analysis of the approximation power of Butterfly-Net to the Fourier representation of input data shows that the error decays exponentially as the depth increases.
1 code implementation • 25 Oct 2018 • Xiuyuan Cheng, Gal Mishne
The extraction of clusters from a dataset which includes multiple clusters and a significant background component is a non-trivial task of practical importance.
1 code implementation • ECCV 2020 • Henry Li, Ofir Lindenbaum, Xiuyuan Cheng, Alexander Cloninger
Variational autoencoders (VAEs) and generative adversarial networks (GANs) enjoy an intuitive connection to manifold learning: in training the decoder/generator is optimized to approximate a homeomorphism between the data distribution and the sampling space.
no code implementations • 24 Sep 2019 • Wei Zhu, Qiang Qiu, Robert Calderbank, Guillermo Sapiro, Xiuyuan Cheng
Encoding the scale information explicitly into the representation learned by a convolutional neural network (CNN) is beneficial for many computer vision tasks especially when dealing with multiscale inputs.
no code implementations • NeurIPS 2020 • Ze Wang, Xiuyuan Cheng, Guillermo Sapiro, Qiang Qiu
In this paper, we consider domain-invariant deep learning by explicitly modeling domain shifts with only a small amount of domain-specific parameters in a Convolutional Neural Network (CNN).
1 code implementation • 25 Sep 2019 • Xiuyuan Cheng, Alexander Cloninger
The recent success of generative adversarial networks and variational learning suggests training a classifier network may work well in addressing the classical two-sample problem.
no code implementations • 25 Sep 2019 • Ze Wang, Xiuyuan Cheng, Guillermo Sapiro, Qiang Qiu
Domain shifts are frequently encountered in real-world scenarios.
no code implementations • ICLR 2020 • Ze Wang, Xiuyuan Cheng, Guillermo Sapiro, Qiang Qiu
One of these questions is how to efficiently achieve proper diversity and sampling of the multi-mode data space.
no code implementations • 25 Sep 2019 • Wei Zhu, Qiang Qiu, Robert Calderbank, Guillermo Sapiro, Xiuyuan Cheng
Encoding the input scale information explicitly into the representation learned by a convolutional neural network (CNN) is beneficial for many vision tasks especially when dealing with multiscale input signals.
1 code implementation • 9 Dec 2019 • Zhongshu Xu, Yingzhou Li, Xiuyuan Cheng
Structured CNN designed using the prior information of problems potentially improves efficiency over conventional CNNs in various tasks in solving PDEs and inverse problems in signal processing.
2 code implementations • ICLR 2021 • Xiuyuan Cheng, Zichen Miao, Qiang Qiu
Recent deep models using graph convolutions provide an appropriate framework to handle such non-Euclidean data, but many of them, particularly those based on global graph Laplacians, lack expressiveness to capture local features required for representation of signals lying on the non-Euclidean grid.
no code implementations • 4 Sep 2020 • Ze Wang, Xiuyuan Cheng, Guillermo Sapiro, Qiang Qiu
We then explicitly regularize CNN kernels by enforcing decomposed coefficients to be shared across sub-structures, while leaving each sub-structure only its own dictionary atoms, a few hundreds of parameters typically, which leads to dramatic model reductions.
no code implementations • 3 Nov 2020 • Xiuyuan Cheng, Hau-Tieng Wu
This paper proves the convergence of graph Laplacian operator $L_N$ to manifold (weighted-)Laplacian for a new family of kNN self-tuned kernels $W^{(\alpha)}_{ij} = k_0( \frac{ \| x_i - x_j \|^2}{ \epsilon \hat{\rho}(x_i) \hat{\rho}(x_j)})/\hat{\rho}(x_i)^\alpha \hat{\rho}(x_j)^\alpha$, where $\hat{\rho}$ is the estimated bandwidth function {by kNN}, and the limiting operator is also parametrized by $\alpha$.
1 code implementation • 25 Jan 2021 • Xiuyuan Cheng, Nan Wu
The result holds for un-normalized and random-walk graph Laplacians when data are uniformly sampled on the manifold, as well as the density-corrected graph Laplacian (where the affinity matrix is normalized by the degree matrix from both sides) with non-uniformly sampled data.
no code implementations • 28 Feb 2021 • Yixing Zhang, Xiuyuan Cheng, Galen Reeves
The Gaussian-smoothed optimal transport (GOT) framework, recently proposed by Goldfeld et al., scales to high dimensions in estimation and provides an alternative to entropy regularization.
1 code implementation • 7 May 2021 • Xiuyuan Cheng, Yao Xie
Specifically, when data densities $p$ and $q$ are supported on a $d$-dimensional sub-manifold ${M}$ embedded in an $m$-dimensional space and are H\"older with order $\beta$ (up to 2) on ${M}$, we prove a guarantee of the test power for finite sample size $n$ that exceeds a threshold depending on $d$, $\beta$, and $\Delta_2$ the squared $L^2$-divergence between $p$ and $q$ on the manifold, and with a properly chosen kernel bandwidth $\gamma$.
1 code implementation • NeurIPS 2021 • Xiuyuan Cheng, Yao Xie
We present a novel neural network Maximum Mean Discrepancy (MMD) statistic by identifying a new connection between neural tangent kernel (NTK) and MMD.
1 code implementation • ICLR 2022 • Shixiang Zhu, Haoyun Wang, Zheng Dong, Xiuyuan Cheng, Yao Xie
In this paper, we introduce a novel and general neural network-based non-stationary influence kernel with high expressiveness for handling complex discrete events data while providing theoretical performance guarantees.
no code implementations • 29 Sep 2021 • Ze Wang, Xiuyuan Cheng, Guillermo Sapiro, Qiang Qiu
In other words, a CNN is now reduced to layers of filter atoms, typically a few hundred of parameters per layer, with a common block of subspace coefficients shared across layers.
no code implementations • NeurIPS 2021 • Zichen Miao, Ze Wang, Xiuyuan Cheng, Qiang Qiu
In this paper, we introduce spatiotemporal joint filter decomposition to decouple spatial and temporal learning, while preserving spatiotemporal dependency in a video.
no code implementations • 8 Feb 2022 • Jonathan Zhou, Sarah Huestis-Mitchell, Xiuyuan Cheng, Yao Xie
We present a method to capture groupings of similar calls and determine their relative spatial distribution from a collection of crime record narratives.
1 code implementation • 17 Feb 2022 • Chen Xu, Xiuyuan Cheng, Yao Xie
We propose an alternative approach to neural network training using the monotone vector field, an idea inspired by the seminal work of Juditsky and Nemirovski [Juditsky & Nemirovsky, 2019] developed originally to solve parameter estimation problems for generalized linear models (GLM) by reducing the original non-convex problem to a convex problem of solving a monotone variational inequality (VI).
1 code implementation • 2 Jun 2022 • Chen Xu, Xiuyuan Cheng, Yao Xie
The proposed model consists of an invertible sub-network that maps one-to-one from data to an intermediate encoded feature, which allows forward prediction by a linear classification sub-network as well as efficient generation from output labels via a parametric mixture model.
1 code implementation • 14 Jun 2022 • Ziyu Chen, Yingzhou Li, Xiuyuan Cheng
The current paper introduces a new neural network approach, named SpecNet2, to compute spectral embedding which optimizes an equivalent objective of the eigen-problem and removes the orthogonalization layer in SpecNet1.
1 code implementation • 22 Jun 2022 • Xiuyuan Cheng, Boris Landa
This paper proves the convergence of bi-stochastically normalized graph Laplacian to manifold (weighted-)Laplacian with rates, when $n$ data points are i. i. d.
1 code implementation • 7 Jul 2022 • Matthew Repasky, Xiuyuan Cheng, Yao Xie
In this paper, we investigate the role of $L^2$ regularization in training a neural network Stein critic so as to distinguish between data sampled from an unknown probability distribution and a nominal model distribution.
no code implementations • 16 Sep 2022 • Boris Landa, Xiuyuan Cheng
The Gaussian kernel and its traditional normalizations (e. g., row-stochastic) are popular approaches for assessing similarities between data points.
no code implementations • 31 Oct 2022 • Tingnan Gong, Junghwan Lee, Xiuyuan Cheng, Yao Xie
Change-point detection, detecting an abrupt change in the data distribution from sequential data, is a fundamental problem in statistics and machine learning.
no code implementations • 21 Nov 2022 • Zheng Dong, Xiuyuan Cheng, Yao Xie
Another popular type of deep model for point process data is based on representing the influence kernel (rather than the intensity function) by neural networks.
1 code implementation • NeurIPS 2023 • Chen Xu, Xiuyuan Cheng, Yao Xie
Normalizing flow is a class of deep generative models for efficient sampling and likelihood estimation, which achieves attractive performance, particularly in high dimensions.
no code implementations • 29 Mar 2023 • Eitan Rosen, Paulina Hoyos, Xiuyuan Cheng, Joe Kileel, Yoel Shkolnisky
We introduce the G-invariant graph Laplacian that generalizes the graph Laplacian by accounting for the action of the group on the data set.
no code implementations • 19 May 2023 • Chen Xu, Xiuyuan Cheng, Yao Xie
Flow-based models are widely used in generative tasks, including normalizing flow, where a neural network transports from a data distribution $P$ to a normal distribution.
1 code implementation • 2 Jun 2023 • Yixuan Tan, Liyan Xie, Xiuyuan Cheng
We propose an RNN-based model, called RNN-ODE-Adap, that uses a neural ODE to represent the time development of the hidden states, and we adaptively select time steps based on the steepness of changes of the data over time so as to train the model more efficiently for the "spike-like" time series.
no code implementations • 12 Jun 2023 • Eitan Rosen, Xiuyuan Cheng, Yoel Shkolnisky
The diffusion maps embedding of data lying on a manifold have shown success in tasks ranging from dimensionality reduction and clustering, to data visualization.
no code implementations • 20 Jun 2023 • Zheng Dong, Matthew Repasky, Xiuyuan Cheng, Yao Xie
Point process models are widely used for continuous asynchronous event data, where each data point includes time and additional information called "marks", which can be locations, nodes, or event types.
no code implementations • 26 Oct 2023 • Xiuyuan Cheng, Jianfeng Lu, Yixin Tan, Yao Xie
Flow-based generative models enjoy certain advantages in computing the data generation and the likelihood, and have recently shown competitive empirical performance.
1 code implementation • 30 Oct 2023 • Chen Xu, JongHyeok Lee, Xiuyuan Cheng, Yao Xie
We present a computationally efficient framework, called $\texttt{FlowDRO}$, for solving flow-based distributionally robust optimization (DRO) problems with Wasserstein uncertainty sets while aiming to find continuous worst-case distribution (also called the Least Favorable Distribution, LFD) and sample from it.