Empirical study shows standard CNNs deviate from NTK predictions.
problem Understanding how standard finite-width CNNs behave compared to their infinite-width NTK counterparts.
method Empirical analysis of AlexNet and LeNet architectures.
result Standard CNNs deviate significantly from their NTK counterparts, but deviation decreases with wider networks.
Wide CNNs outperform infinite width networks, revealing scaling laws.
problem Understanding the performance difference between finite and infinite width convolutional networks.
method Diagrammatic approach to derive asymptotic width dependence for various quantities.
result The difference in performance between finite and infinite width models vanishes at a definite rate with respect to model width.
Convolutional neural networks handle rotated image symmetries without dimensionality issues.
problem Binary image classification with rotational symmetry.
method Least squares plug-in classifiers based on convolutional neural networks under rotationally symmetric assumptions.
result Convolutional neural networks can circumvent the curse of dimensionality in binary image classification with rotational symmetry.
This paper removes the finite variance assumption for deep convolutional neural networks.
problem Removing the finite variance assumption for deep convolutional neural networks.
method Assuming iid parameters distributed according to a stable distribution, the paper shows that the infinite-channel limit of a deep feed-forward convolutional neural network is a multivariate stable stochastic process.
result The infinite-channel limit of a deep feed-forward convolutional neural network, under suitable scaling, is a multivariate stable stochastic process.
We establish large deviation principles for convolutional neural networks.
problem Understanding the behavior of convolutional neural networks in the infinite-channel limit.
method We establish large deviation principles for convolutional neural networks under Gaussian prior and posterior distributions.
result We provide a large deviation principle for the sequence of conditional covariance matrices and the posterior distribution.
Paper presents efficient algorithms for convolutional neural networks using Winograd minimal filtering.
problem Resource-efficient implementation of convolutional neural networks.
method Winograd minimal filtering trick applied to M-tap filters (M=3,5,7,9,11) for parallel hardware implementation.
result Approximately 30% reduction in multipliers for fully parallel hardware implementation.
We present a novel class of convolutional neural networks (CNNs) for set functions, i.e., data indexed with the powerset of a finite set. The convolutions are derived as linear, shift-equivariant functions for various notions of shifts on set functions. The framework is fundamentally different from graph convolutions b…
Neural network outperforms traditional methods in chaotic dynamics classification.
problem Classifying chaotic and regular dynamics of the Chirikov standard map.
method Trained a convolutional neural network on finite-length trajectories compared to traditional Lyapunov exponent computation.
result Neural network outperforms traditional methods for short periods, converging faster and more robustly.
New image classifier uses hierarchical max-pooling with local pooling.
problem Improving image classification accuracy with variable spatial relationships.
method Introduces a hierarchical max-pooling model with additional local pooling for convolutional neural networks.
result Demonstrates improved performance in estimating image features.
FiniteNet uses a neural network to improve PDE solving methods.
problem Improving accuracy in solving time-dependent PDEs.
method Fully convolutional LSTM network trained on simulation data.
result Reduces error by a factor of 2 to 3 compared to baseline methods.
We study how finite Bayesian neural networks adapt their hidden representations.
problem Understanding how finite Bayesian neural networks differ from infinite ones.
method We analyze the asymptotics of learned feature kernels for various network architectures.
result The leading finite-width corrections to feature kernels have a universal form.
Analyzes DNNs trained with noisy gradients, finding FWCs negligible for large n.
problem Analyzing DNNs trained with noisy gradients.
method Introduced analytical framework to analyze non-Gaussian stochastic process.
result FWCs negligible for large n, improving CNN performance.
Theory captures feature learning effects in finite CNNs.
problem Feature learning in finite deep neural networks.
method Derive a self-consistent Gaussian Process theory.
result Good agreement with experiments and sharp transition between regimes.
Ensembles of neural networks improve training dynamics and performance.
problem Improving neural network performance through model size increase.
method Defining collegial ensembles (CE) as multiple independent models trained as a single model, and using theoretical results on NTK to optimize architecture search.
result CE dynamics simplify and scale favorably, resembling wide models, and can be efficiently implemented using group convolutions and block diagonal layers.
Infinite CNNs lose spatial correlations, but can be restored by correlated weights.
problem Infinite CNNs lose spatial correlations, which are crucial for their performance.
method Introduced correlated weights to restore spatial correlations in infinite CNNs.
result Optimal performance is achieved with a moderate level of weight correlation.
VC dimensions of group CNNs are infinite for certain kernels and groups.
problem Estimating the generalization capacity of group convolutional neural networks.
method Identifying precise VC dimension estimates for simple sets of group CNNs.
result Two-parameter families of convolutional neural networks have an infinite VC dimension for infinite groups and certain kernels.
Optimal rates for shallow ReLU networks in nonparametric regression.
problem Approximating smooth and non-smooth functions with shallow ReLU networks.
method Analysis of shallow ReLUk neural networks, using variation norms and deep learning theory. result Optimal approximation rates for shallow ReLU networks in nonparametric regression.
Convolutional Neural Networks, as most artificial neural networks, are commonly viewed as methods different in essence from kernel-based methods. We provide a systematic translation of Convolutional Neural Networks (ConvNets) into their kernel-based counterparts, Convolutional Kernel Networks (CKNs), and demonstrate th…
Empirical study compares wide neural networks to kernel methods, resolving open questions.
problem Understanding the relationship between wide neural networks and kernel methods.
method Large-scale empirical study using various neural network architectures and kernel methods.
result Wide neural networks outperform fully-connected finite-width networks in some cases, but underperform convolutional finite-width networks.
Convolutional neural networks converge quickly with gradient descent.
problem Learning efficient image classifiers with over-parameterized networks.
method Gradient descent for training over-parametrized CNNs with global average-pooling.
result Gradient descent quickly reduces the misclassification risk of CNNs.
Graph convolutional deep kernel machine learns representations for graph tasks.
problem Limited representation learning in infinite-width neural networks.
method Developed a graph convolutional deep kernel machine as an infinite-width limit.
result Representation learning improves performance for heterophilous node classification tasks.
Convolutional neural networks improve image classification accuracy.
problem Improving accuracy in image classification.
method Analyzing the convergence rate of misclassification risk for image classifiers.
result A rate of convergence independent of image dimension proves the effectiveness of CNNs.
Deep neural network predicts traffic flow on city maps.
problem Short-term traffic flow prediction on high-resolution city maps.
method UNet-based deep convolutional neural network with densely connected layers.
result Best performance on the Traffic4cast challenge 2019.
Framework for differentiating WFSTs for structured loss functions.
problem Training dynamic structured loss functions in neural networks.
method Automatic differentiation with WFSTs, combining pruning and back-off.
result Demonstrated learning over WFST latent phrase decomposition.
We applied pre-defined kernels also known as filters or masks developed for image processing to convolution neural network. Instead of letting neural networks find its own kernels, we used 41 different general-purpose kernels of blurring, edge detecting, sharpening, discrete cosine transformation, etc. for the first la…
Learning representations that accurately capture long-range dependencies in sequential inputs -- including text, audio, and genomic data -- is a key problem in deep learning. Feed-forward convolutional models capture only feature interactions within finite receptive fields while recurrent architectures can be slow and …
This work proposes hyperbolic deep convolutional neural networks for better pattern recognition.
problem The limitations of Euclidean deep convolutional neural networks in capturing intricate patterns.
method Developed Hyperbolic DCNN based on Poincaré Disc, analyzing expansive convolution in non-Euclidean space.
result Hyperbolic convolutional architecture outperforms Euclidean ones in pattern recognition tasks.
Proves DCNNs with expansive convolution are strongly universally consistent.
problem Theoretical consistency of deep convolutional neural networks (DCNNs).
method Empirical risk minimization on DCNNs with expansive convolution (with zero-padding).
result DCNNs with expansive convolution are strongly universally consistent.
This research studies affine invariance in continuous-domain convolutional neural networks.
problem Recognizing patterns and features under affine transformations in continuous domains.
method Introduces a new criterion for assessing affine invariance, embeds images into the affine Lie group, and analyzes convolution over this group.
result Extends the scope of geometrical transformations that deep-learning pipelines can handle.
L-CNNs learn gauge invariant quantities on lattices.
problem Learning gauge invariant quantities on lattices.
method Novel convolutional layer preserving gauge equivariance and forming Wilson loops.
result L-CNNs can approximate any gauge covariant function on the lattice.
Convolutional neural networks (CNNs) have achieved breakthrough performances in a wide range of applications including image classification, semantic segmentation, and object detection. Previous research on characterizing the generalization ability of neural networks mostly focuses on fully connected neural networks (F…
L-CNNs maintain gauge symmetry on non-Abelian lattice theories.
problem Applying convolutional neural networks to non-Abelian lattice gauge theories while preserving gauge symmetry.
method Developed a geometric formulation of L-CNNs that are equivariant under global symmetries and gauge transformations.
result Convolutional operations in L-CNNs are a specific case of gauge-equivariant neural networks on SU(N) principal bundles. Convolutional neural network is an important model in deep learning. To avoid exploding/vanishing gradient problems and to improve the generalizability of a neural network, it is desirable to have a convolution operation that nearly preserves the norm, or to have the singular values of the transformation matrix corresp…
Proposes a new convolutional neural network for non-grid data.
problem Limited applicability of standard CNNs to non-grid structured data.
method Introduces Parametric Continuous Convolution (PCC) with learnable kernel functions.
result Significant improvement in point cloud segmentation and lidar motion estimation.
Recently, graph neural networks have been adopted in a wide variety of applications ranging from relational representations to modeling irregular data domains such as point clouds and social graphs. However, the space of graph neural network architectures remains highly fragmented impeding the development of optimized …
Simplicial neural networks extend graph neural networks to handle higher-order interactions.
problem Handling higher-order interactions in complex data structures.
method Define a convolution operation for simplicial complexes and use it to construct convolutional neural networks.
result SNNs effectively impute missing data in coauthorship complexes.
New neural model processes 2D data with long-range dependencies efficiently.
problem Limited receptive field of convolutions for complex 2D tasks.
method Proposes Matrix Shuffle-Exchange network with O(logn) layers and O(n2logn) complexity. result Exceeds convolutional and graph neural network baselines in long-range dependency modeling.
We introduce Group equivariant Convolutional Neural Networks (G-CNNs), a natural generalization of convolutional neural networks that reduces sample complexity by exploiting symmetries. G-CNNs use G-convolutions, a new type of layer that enjoys a substantially higher degree of weight sharing than regular convolution la…
Convolutional networks struggle with repeating patterns in ECGs.
problem Modeling repeating patterns in electrocardiogram signals.
method Demonstrated through ECG examples, highlighting systemic issues in deep learning.
result Counterintuitive effects on generalization in deep networks.
The study examines deep convolutional neural networks and their learning ability.
problem Understanding the learning ability of deep convolutional neural networks (DCNNs).
method Examines DCNNs under both underparameterized and overparameterized settings, using a novel network deepening scheme.
result Establishes the first learning rates of underparameterized DCNNs and shows how adding layers can create interpolating DCNNs with good learning rates.
Combines BERT and graph CNN for improved text classification.
problem Text classification problems
method Combining BERT embedding and graph convolutional neural network
result Graph CNN model performs better than classical models combined with BERT
Deep learning has been widely applied and brought breakthroughs in speech recognition, computer vision, and many other domains. The involved deep neural network architectures and computational issues have been well studied in machine learning. But there lacks a theoretical foundation for understanding the approximation…
A new CNN approach for time series forecasting outperforms traditional RNNs.
problem Time series forecasting using conventional RNNs.
method Temporally folded convolutional neural networks (TFC's) for sequence forecasting.
result TFC's outperform conventional RNNs on sequential MNIST and JSB chorals datasets.
In this paper, some patterns of the Neuron Response of deep Convolutional Neural Networks were observed.
Convolutional neural networks predict the analytic rank of elliptic curves accurately.
problem Predicting the analytic rank of elliptic curves over Q.
method Applied one-dimensional convolutional neural networks to Frobenius traces.
result High accuracy predictions for analytic rank across various conductors.
Study deep maxout networks and their equivalence to Gaussian processes.
problem Understanding neural networks with infinite width.
method Derive equivalence between deep maxout networks and Gaussian processes, characterize maxout kernel, and provide efficient numerical implementation.
result Bayesian inference based on deep maxout network kernel leads to competitive results compared to finite-width counterparts and deep neural network kernels.
This paper reviews graph convolutional neural networks (GCNNs) through the lens of edge-variant graph filters. The edge-variant graph filter is a finite order, linear, and local recursion that allows each node, in each iteration, to weigh differently the information of its neighbors. By exploiting this recursion, we fo…
New method tightens Lipschitz bounds for CNNs efficiently.
problem Lipschitz regularization of Convolutional Neural Networks (CNNs).
method Using Toeplitz matrix theory, introduces a tight and computationally efficient upper bound for convolutional layers.
result Developed an algorithm to train Lipschitz regularized CNNs.