Differentially-private FNAS protects privacy while collaboratively searching for neural architectures.
problem Collaborative neural architecture search with privacy concerns.
method Federated Neural Architecture Search (FNAS) with differential privacy (DP-FNAS).
result DP-FNAS can search for highly-performant neural architectures while protecting individual parties' privacy.
Proposes a method to improve neural architectures reproducibly.
problem Lack of reproducibility in Neural Architecture Transformer (NAT).
method Differentiable Neural Architecture Transformation (DNAT).
result DNAT outperforms NAT and is applicable to various models and datasets.
New model predicts neural network performance from early training epochs, incorporating architecture impact.
problem Predicting neural network performance from early training epochs, neglecting architecture impact.
method Architecture-aware graph ordinary differential equation model.
result Model outperforms state-of-the-art methods for MLP and CNN learning curves.
DNArch learns CNN architectures by backpropagation.
problem Discovering optimal CNN architectures.
method Differentiable Neural Architectures (DNArch) learns CNN architectures by backpropagation, controlling kernel sizes, channels, downsampling positions, and depth.
result DNArch finds performant CNN architectures across various tasks.
Efficient neural networks compute various differential operators cheaply.
problem Efficient computation of higher time complexity differential operators.
method Restricted neural network architectures with diagonal and hollow Jacobian matrices, allowing efficient extraction of dimension-wise derivatives.
result Demonstrated efficient computation of differential operators for various applications.
HardCoRe-NAS finds fitting neural networks adhering to hard resource constraints.
problem Finding fitting neural networks that adhere to hard resource constraints.
method Accurate formulation of resource requirement and scalable search method.
result HardCoRe-NAS generates state-of-the-art architectures strictly satisfying hard resource constraints.
Differentiable NAS frameworks grow networks wider and deeper, revealing biases in wiring evolution.
problem Understanding the evolution of neural architecture wiring in differentiable NAS methods.
method Unified view on searching algorithms, local cost minimization, empirical and theoretical analyses.
result Implicit inductive biases cause observed searching patterns in differentiable NAS methods.
New neural stack and Turing Machine architectures prove stability and computational power.
problem Designing stable neural network architectures for Turing Machine simulation.
method Introducing neural stack and Turing Machine architectures, proving stability and computational equivalence.
result Differentiable nnTM with bounded neurons can simulate Turing Machine in real-time and is equivalent to UTM.
New deep learning architecture learns martingales efficiently.
problem Efficiently learning martingales in financial derivatives pricing.
method High-order weak approximation algorithms of Runge-Kutta type.
result Deep neural networks based on this architecture learn martingales effectively.
UNAS combines DNAS and RL for efficient architecture search.
problem Discovering high accuracy or low latency neural architectures.
method Unified framework combining differentiable and reinforcement learning approaches.
result UNAS achieves state-of-the-art accuracy on CIFAR-10, CIFAR-100, and ImageNet datasets.
DANCE optimizes neural network and accelerator design for faster, more efficient DNN execution.
problem Challenges in optimizing neural network and accelerator design for efficient DNN execution.
method Differentiable approach to co-exploration of accelerator and network architecture design.
result Significantly shorter time to achieve superior accuracy and hardware cost metrics.
FiGS searches over a larger space of architectures for efficient mobile models.
problem Designing small, efficient deep networks for mobile devices.
method Differentiable search method using sparse regularization and Logistic-Sigmoid distribution.
result FiGS produces state-of-the-art parameter-efficient models on ImageNet and improves object detection performance.
PINNs solve differential geometry problems in complex shapes.
problem Solving differential geometry problems in complex shapes.
method Training neural networks with loss functions inspired by differential conditions.
result PINNs are effective for differential geometry problems.
Paper presents an ADMM-based approach to efficiently integrate quadratic programming layers into neural networks.
problem Integrating quadratic programs into neural networks for optimization.
method An ADMM-based network layer architecture for solving quadratic programs efficiently.
result The ADMM layer is approximately an order of magnitude faster than existing methods for medium scaled problems.
Paper uses CMAB to improve NAS efficiency and accuracy.
problem Improving efficiency and accuracy of NAS for DNNs.
method Formulated NAS as CMAB, used Nested Monte-Carlo Search.
result Discovered cell structure achieves comparable accuracy to state-of-the-art, 20x faster.
We propose Stochastic Neural Architecture Search (SNAS), an economical end-to-end solution to Neural Architecture Search (NAS) that trains neural operation parameters and architecture distribution parameters in same round of back-propagation, while maintaining the completeness and differentiability of the NAS pipeline.…
Bonsai-Net efficiently discovers state-of-the-art models with fewer parameters.
problem Efficiently discovering state-of-the-art neural architectures with minimal computational expense.
method Bonsai-Net uses a modified differential pruner to explore a relaxed search space.
result Bonsai-Net consistently discovers better architectures than random search with fewer parameters.
Proposes a new method to optimize graph neural network architectures on heterogeneous information networks.
problem Weaknesses in instability and inflexibility of existing graph neural architecture search methods.
method Partial Message Meta Multigraph search (PMMM) using a differentiable framework to search for a meaningful meta multigraph.
result Significantly more stable and effective than state-of-the-art heterogeneous GNNs.
DrNAS improves neural architecture search with Dirichlet distribution and progressive learning.
problem Efficiently search for neural architectures with improved generalization and exploration.
method Formulates architecture search as a distribution learning problem using Dirichlet distribution and gradient-based optimization. Introduces a progressive learning scheme to handle large-scale tasks.
result Achieves state-of-the-art results on CIFAR-10 and ImageNet, demonstrating improved generalization and exploration.
Paper tackles unfair advantages in DARTS, presenting Fair DARTS to improve neural architecture search.
problem Performance collapse in DARTS due to unfair advantages in skip connections.
method Relax exclusive competition to collaborative, let architectural weights be independent, and use zero-one loss for discretization.
result New state-of-the-art results on CIFAR-10 and ImageNet, demonstrating the effectiveness of Fair DARTS.
NDM incorporates geometric structure into neural networks for better optimization and interpretability.
problem Efficient and interpretable deep learning architectures.
method NDM is a neural network architecture that explicitly incorporates geometric structure into its design, using a Coordinate Layer, Geometric Layer, and Evolution Layer.
result NDM provides intrinsic regularization, enhancing generalization and robustness.
We provide a proof of backpropagation algorithm in matrix notation.
problem The lack of a full induction proof of backpropagation algorithm in matrix notation.
method We provide a full induction proof of the BP algorithm in matrix notation, situating it in the framework of matrix differential calculus.
result We prove the validity of the backpropagation algorithm in inductive form.
Neural architecture search (NAS) recently attracts much research attention because of its ability to identify better architectures than handcrafted ones. However, many NAS methods, which optimize the search process in a discrete search space, need many GPU days for convergence. Recently, DARTS, which constructs a diffe…
This work proposes searching for optimal operation distribution in neural architecture search.
problem Finding optimal neural architecture with specific operations and connections.
method Search for the optimal operation distribution, providing a stochastic and approximate solution.
result Operation distribution holds enough discriminating power to reliably identify a solution and is easier to optimise than traditional encodings.
Novel NAS method balances performance and hardware metrics efficiently.
problem Challenging multi-objective optimization in neural architecture search.
method Parameterizes joint architectural distribution via hypernetwork conditioned on hardware features and preferences.
result Zero-shot transferability to new devices with representative and diverse architectures.
PDNAS optimizes GNN architectures for diverse datasets.
problem Inadequate adaptability and combinatorial search space in GNNs.
method Dual architecture search (micro- and macro-architectures) with gradient-based optimization.
result PDNAS finds deeper GNNs with better performance on diverse datasets.
Extends NAS to learn both intra-cell and inter-cell architectures for language modeling.
problem Limited NAS systems restrict search to recurrent or convolutional cells.
method Designs a joint learning method to perform intra-cell and inter-cell NAS simultaneously.
result Significantly outperforms a strong baseline on PTB and WikiText data.
CRUs model irregular time series with continuous hidden states.
problem Handling irregular time intervals in sequential data.
method Continuous Recurrent Units (CRUs) that integrate hidden states via a linear stochastic differential equation.
result CRUs outperform methods based on neural ordinary differential equations in irregular time series interpolation.
New ADANNs improve PDE approximations.
problem Approximating operators for parametric PDEs.
method Custom ANN architectures and initialization schemes.
result ADANNs significantly outperform existing methods.
New framework combines simple machines into complex ones for better neural network performance.
problem Improving neural network performance with limited training data.
method Developed a framework using topology and functional analysis to combine simple machines into complex ones, and used kernel methods to find optimal architectures.
result Kernel-inspired networks can outperform classical neural networks when training data is small.
INNs can approximate diverse functions despite layer restrictions.
problem Can INNs approximate sufficiently diverse functions?
method Developed a theoretical framework based on differential geometry to simplify the approximation problem of diffeomorphisms.
result INNs have the universal approximation property.
New framework explains neural network bias in solving differential equations.
problem Understanding and controlling the bias in PINNs for differential equations.
method Deriving an integro-differential equation from PINNs and GPR equivalence.
result PINN predictions are influenced by a kernel term reflecting architecture choices.
Neural architecture search (NAS) has a great impact by automatically designing effective neural network architectures. However, the prohibitive computational demand of conventional NAS algorithms (e.g. 104 GPU hours) makes it difficult to \emph{directly} search the architectures on large-scale tasks (e.g. ImageNet).…
FNN approximates functions and solves PDEs with periodic BCs.
problem Approximating and solving periodic functions and PDEs.
method Fourier neural network architecture with activation and loss functions.
result FNN can solve PDEs with periodic BCs and is interpretable.
Smooth embedding space improves NAS performance.
problem Efficiently predicting good neural architectures.
method Two-sided variational graph autoencoder.
result Smooth embedding space facilitates extrapolation to unseen architectures.
In many learning situations, resources at inference time are significantly more constrained than resources at training time. This paper studies a general paradigm, called Differentiable ARchitecture Compression (DARC), that combines model compression and architecture search to learn models that are resource-efficient a…
In our work, we bridge deep neural network design with numerical differential equations. We show that many effective networks, such as ResNet, PolyNet, FractalNet and RevNet, can be interpreted as different numerical discretizations of differential equations. This finding brings us a brand new perspective on the design…
DSNAS optimizes neural architecture and parameters in one step.
problem Poor correlation between architecture performance in two stages of NAS methods.
method Task-specific end-to-end approach with DSNAS framework.
result DSNAS discovers comparable accuracy networks in less time.
NoisyDARTS injects random noise to improve neural architecture search.
problem Performance collapse in Differentiable Architecture Search (DARTS).
method Inject unbiased random noise to skip connections to impede gradient flow.
result NoisyDARTS achieves state-of-the-art results across various tasks.
A neural program synthesis method with iterative fix operations.
problem Creating correct programs from input-output examples.
method Combines encoder-decoder synthesis with a differentiable fixer.
result Improves synthesis accuracy by reducing discrepancies between outputs and desired outputs.
Paper discovers differential equations from data using neural networks and Bayesian methods.
problem Discovering differential equations from datasets using machine learning.
method Integrates neural network-based surrogates with Sparse Bayesian Learning (SBL).
result Proposes a robust model discovery algorithm and a Physics Informed Normalizing Flow (PINF).
Graph neural controlled differential equations learn graph dynamics from vertex observations.
problem Predicting future states of dynamical systems on graphs with limited vertex data.
method Incorporates graph topology information into NCDE to predict graph dynamics.
result Informed NCDE requires fewer parameters and lower MAE compared to previous methods.
New method learns vector fields from noisy time series data.
problem Learning vector fields from noisy time series data.
method Neural network architecture with tensor products of one-dimensional neural shape functions for vector field approximation, alternating minimization for noise handling.
result Neural shape function architecture robust to noise, learning accurate vector fields from data with up to 10% Gaussian noise.
To understand the fundamental trade-offs between training stability, temporal dynamics and architectural complexity of recurrent neural networks~(RNNs), we directly analyze RNN architectures using numerical methods of ordinary differential equations~(ODEs). We define a general family of RNNs--the ODERNNs--by relating t…
VINNAS uses variational inference to avoid mode collapse in neural architecture search.
problem Mode collapse in gradient-based NAS methods, leading to suboptimal architectures.
method Differentiable variational inference with variational dropout and automatic relevance determination.
result State-of-the-art accuracy with up to twice fewer non-zero parameters.
DP-BNNs improve accuracy, privacy, and reliability in neural networks.
problem Balancing privacy and accuracy in neural networks.
method Proposed three DP-BNNs: DP-SGLD, DP-BBP, and DP-MC Dropout.
result DP-SGLD achieves high accuracy under strong privacy guarantees.
This paper presents OptNet, a network architecture that integrates optimization problems (here, specifically in the form of quadratic programs) as individual layers in larger end-to-end trainable deep networks. These layers encode constraints and complex dependencies between the hidden states that traditional convoluti…
This paper compares deeper and wider neural networks for optimal generalization error in Sobolev losses.
problem The dilemma of choosing between deeper or wider neural networks for optimal generalization error.
method Analytical investigations into the influence of sample points, parameters, and loss function regularity on neural network architecture.
result A higher number of parameters favors wider neural networks, while more sample points and greater loss function regularity favor deeper neural networks.