SPTN uses invertible transformations to improve sum-product networks.
problem Improving inference efficiency and tractability in sum-product networks.
method Integrates invertible transformations into sum-product networks (SPNs).
result SPTNs with Gaussian leaves and affine transformations are as tractable as SPNs.
Planar neural networks learn image transformations from sequences.
problem Learning image transformations for mental simulation.
method Using planar neural networks, the study investigates various factors affecting the learning of image transformations.
result The approach can effectively learn and transfer image transformations, including translation, rotation, and scaling.
Adversarial learning improves image augmentation for neural networks.
problem Improving data augmentation for neural networks with limited data.
method Adversarial learning using an encoder-decoder architecture with a spatial transformer network.
result Our approach outperforms previous generative data augmentation methods.
Improved wavelet filters enhance neural network performance.
problem Enhancing wavelet transform for better neural network representations.
method Extending gradient-based filter learning to dual-tree wavelet transform.
result Directional filters improve dual-tree wavelet transform performance.
Transformers can simulate MLE for Bayesian network sequences.
problem Understanding transformers' capabilities in Bayesian network sequence generation.
method In-context maximum likelihood estimation (MLE) for autoregressive sequence generation.
result A simple transformer model can estimate Bayesian network probabilities and generate new samples.
Researchers improve transformer networks' optimization and understanding.
problem Improving the understanding and optimization of transformer networks.
method Introducing a convex alternative to the self-attention mechanism and reformulating the training problem as a convex optimization problem.
result Revealed an implicit regularization mechanism that promotes sparsity across tokens.
Unified method for deriving ridgelet transforms for various neural network architectures.
problem Deriving closed-form expressions for ridgelet transforms in modern neural network architectures.
method Unified Fourier slice method to derive ridgelet transforms for diverse neural network types.
result Systematic method to derive ridgelet transforms for various neural network architectures.
Graph scattering transforms are stable to metric perturbations of network topology.
problem Stability of graph data representations under metric perturbations.
method Extending scattering transforms to network data using multiresolution graph wavelets and graph convolutions.
result Graph scattering transforms are stable to metric perturbations of the underlying network topology.
CNNs encode data augmentation transformations, especially in early layers.
problem Whether neural network features encode data augmentation transformations.
method Systematic approach using pre-trained vision models to predict augmentation transformations.
result Neural network features encode data augmentation transformations, especially in early layers.
B-cos transforms improve neural network interpretability by aligning weights.
problem Improving interpretability of deep neural networks.
method Replacing linear transforms with B-cos transforms that promote weight-input alignment during training.
result B-cos transforms lead to highly interpretable and task-relevant linear summaries of neural network computations.
Network recasting transforms network architecture for faster inference.
problem Accelerate inference process through network transformation.
method Block-wise recasting of source blocks in a teacher network to target blocks in a student network.
result Transforms network architecture while preserving accuracy and reducing inference time.
Quantum Ridgelet Transform speeds up neural network learning.
problem Efficiently finding sparse trainable subnetworks in neural networks.
method Developed a quantum ridgelet transform (QRT) for linear runtime.
result Quantum Ridgelet Transform efficiently finds sparse trainable subnetworks.
New neural network learns relevant transformations in data, improving object recognition.
problem Current equivariant architectures consider all possible transformations, ignoring relevant ones.
method Co-attentive equivariant neural networks that focus on co-occurring transformations.
result Outperforms conventional equivariant networks on rotated MNIST and CIFAR-10.
Unified graph scattering transforms improve theoretical properties of graph neural networks.
problem Improving theoretical guarantees for graph neural networks.
method Introducing windowed and non-windowed geometric scattering transforms for graphs.
result Unified family of graph scattering transforms with provable stability and invariance.
CaTs use DAGs with transformers to enforce causal constraints, improving neural network robustness.
problem Neural networks lack inherent causal structure respect, leading to reliability issues.
method Introducing Causal Transformers (CaTs) that operate under predefined causal constraints specified by DAGs.
result CaTs improve robustness and interpretability of neural networks under causal constraints.
UGConvs improve CNN accuracy with unitary transforms.
problem Improving CNN accuracy with richer representations.
method UGConvs combine group convolutions with unitary transforms.
result HadaNets achieve similar accuracy to circulant networks with lower complexity.
The study analyzes the convergence rate of a large transformer model.
problem Understanding the convergence rate of over-parameterized transformer models.
method Theoretical analysis focusing on gradient descent optimization.
result Theoretical upper bound on missclassification probability.
Proposes a neural network method to correct residual distortions in coordinate transformations.
problem Nonlinear and spatially dependent distortions in coordinate transformation models.
method Residual-based neural network approach focusing on systematic distortions.
result The method improves accuracy and stability in challenging conditions.
Enhances non-life insurance pricing models using transformer models.
problem Improving predictive power of non-life insurance pricing models.
method Enhances actuarial non-life models with transformer models for tabular data.
result Transformer models outperform benchmark models in claim frequency prediction.
New network transformation for efficient architecture search.
problem Limitation of current network transformation operations that can only modify layers, not paths.
method Path-level transformation operations using a bidirectional tree-structured reinforcement learning meta-controller.
result Improved parameter efficiency and better test results (97.70% test accuracy on CIFAR-10).
Transforms between neural networks using manifold-learning techniques.
problem Establish equivalence between different neural networks.
method Diffusion maps with a Mahalanobis-like metric to construct transformations between network outputs and internal neuron activations.
result Established equivalence classes between neural networks trained on various data types.
Top-down information plays a central role in human perception, but plays relatively little role in many current state-of-the-art deep networks, such as Convolutional Neural Networks (CNNs). This work seeks to explore a path by which top-down information can have a direct impact within current deep networks. We explore …
The paper explores how generative networks can transform noise distributions into other distributions.
problem Transforming noise distributions into desired distributions using generative networks.
method Developed a space-filling function for ReLU networks and provided efficient methods for univariate uniform to normal distribution transformations.
result Optimal construction for ReLU networks to increase noise dimensionality and efficient methods for distribution transformations.
We prove that the global minimum of the backpropagation (BP) training problem of neural networks with an arbitrary nonlinear activation is given by the ridgelet transform. A series of computational experiments show that there exists an interesting similarity between the scatter plot of hidden parameters in a shallow ne…
Probabilistic STNs improve image classification and robustness.
problem Training and robustness issues in STNs.
method Probabilistic extension of STNs that estimates stochastic transformations.
result Improved classification performance, robustness, and model calibration.
Develops first robustness verification for complex Transformers.
problem Certify prediction behavior of Transformers with complex self-attention layers.
method Resolves challenges of cross-nonlinearity and cross-position dependency in Transformers.
result Certified robustness bounds are significantly tighter than Interval Bound Propagation.
Transformers can outperform feedforward and recurrent networks due to dynamic sparsity.
problem Understanding when and why Transformers outperform other neural network architectures.
method Analyzing a sequence-to-sequence data generating model with dynamic sparsity, proving sample complexity differences between feedforward, recurrent, and Transformers.
result Transformers can learn dynamic sparsity models with lower sample complexity than feedforward and recurrent networks.
Fusion of transformer networks using optimal transport for improved performance.
problem Improving performance of transformer-based models through fusion.
method Exploiting optimal transport for soft alignment of transformer components.
result Consistently outperforms vanilla fusion and individual parent models.
The key idea of current deep learning methods for dense prediction is to apply a model on a regular patch centered on each pixel to make pixel-wise predictions. These methods are limited in the sense that the patches are determined by network architecture instead of learned from data. In this work, we propose the dense…
Method uses JPEG transform for faster image classification.
problem Efficient image classification with compressed data.
method Reformulates residual networks for JPEG compressed images.
result Mathematically equivalent to spatial domain networks up to ReLu approximation.
GIT-Net uses neural networks to approximate PDE operators efficiently.
problem Approximating PDE operators for complex geometries.
method Parametrizes adaptive generalized integral transforms with deep neural networks.
result GIT-Net outperforms existing neural network operators in multiple areas.
Transformer networks approximate Hölder and Sobolev functions with fixed-depth networks.
problem Nonparametric regression with dependent observations.
method Established novel upper bounds for Transformer networks approximating Hölder and Sobolev functions under various β-mixing data assumptions. result Explicit convergence rates for nonparametric regression problems under β-mixing data assumptions. Neural networks explained through geometric projections.
problem Understanding the geometric and mathematical underpinnings of neural networks.
method Exploiting connections between integration, Radon transforms, and neural networks.
result Distribution of neural network outputs can be interpreted as nonlinear projections along hypersurfaces.
The paper explores properties of the Radon transform in relation to neural networks and ridges.
problem Understanding the Radon transform and its application to neural networks and ridges.
method Investigates properties of the Radon transform, introduces new subspaces, and characterizes ridges for any distributional profile.
result Clarifies and simplifies results on the optimality of ReLU networks using the Radon transform.
Transforms improve CNNs' invariance to image transformations.
problem Current CNN models lack robustness to spatial transformations.
method Randomly transform feature maps during training to learn invariant representations.
result Significant improvements on benchmark tasks, including image recognition and retrieval.
Previous research has shown that computation of convolution in the frequency domain provides a significant speedup versus traditional convolution network implementations. However, this performance increase comes at the expense of repeatedly computing the transform and its inverse in order to apply other network operati…
This thesis explores GNNs, categorizing them into local and global approaches.
problem Understanding the convergence of global GNNs and connecting local and global approaches.
method Categorization of GNNs into local and global, study of Invariant Graph Networks, connecting local and global approaches, and using local MPNN for graph coarsening.
result Established a connection between local and global GNN approaches.
GWNN uses graph wavelets for efficient graph CNNs.
problem Spectral graph CNNs' high computational cost and lack of interpretability.
method Graph wavelet transform for efficient graph convolution.
result GWNN significantly outperforms spectral graph CNNs.
ETs improve model robustness to transformations in images.
problem Improving model robustness to predefined transformations.
method Equivariant Transformers (ETs) incorporating functions equivariant to continuous transformation groups.
result ETs achieve up to 15% relative improvement in error rate on image classification tasks.
The paper introduces models to learn generalized transformation equivariant representations.
problem Capturing intrinsic visual structures equivariant to various transformations.
method Deterministic and probabilistic AutoEncoding Transformations (AET and AVT) models trained to learn visual representations from generic groups of transformations.
result Generalized TERs (GTERs) that are equivariant to transformations in a more general fashion.
Different neural networks learn similar mappings with different weights.
problem Understanding shared representations across neural networks with varying weights.
method Shared response model and orthogonal transformations.
result Different neural networks encode the same input examples as different orthogonal transformations of an underlying shared representation.
The paper provides convergence guarantees for ODE-based generative models using transformers.
problem Theoretical guarantees for ODE-based generative models.
method A pre-trained autoencoder maps inputs to a latent space, and a transformer predicts the velocity field.
result The distribution of samples generated via estimated ODE flow converges to the target distribution in Wasserstein-2 distance.
Variational inference provides approximations to the computationally intractable posterior distribution in Bayesian networks. A prominent medical application of noisy-or Bayesian network is to infer potential diseases given observed symptoms. Previous studies focus on approximating a handful of complicated pathological…
Study presents a method to induce a generalized neural network from joint group invariant functions.
problem Encoding rule of neural network internal data representation.
method Systematic method using joint group invariant function on data-parameter domain.
result Induces a generalized neural network and its inverse operator (ridgelet transform).
Method converts facial expressions and voice of a source speaker into a target speaker.
problem Separate conversion of facial and acoustic features leads to unnatural results.
method Uses three neural networks: conversion, waveform generation, and image reconstruction.
result Significantly higher naturalness achieved when converting both features together.
Neural network models transform physical systems into latent Gaussian distributions.
problem Simplifying and solving classical Hamiltonian systems.
method Symplectic neural networks for canonical transformations.
result Captures nonlinear collective modes in latent space.
Paper improves ATN for generating adversarial examples.
problem Generating adversarial examples to fool models.
method Improves Adversarial Transformation Networks (ATN).
result Won 2nd place in CAAD 2018 non-target task.
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.