This study improves state estimation for nonlinear systems using conditional normalizing flows.
problem Performance degradation of traditional filtering algorithms in nonlinear systems with non-Gaussian uncertainty.
method Uses conditional normalizing flows with MLP, transformer, or state-space models for state and parameter estimation.
result Optimal-transport-inspired kinetic loss mitigates overparameterization in flows.
Matched filters reveal optimal normalization methods for different market participants.
problem Optimizing signal extraction from order flow for market microstructure analysis.
method General matched filter principle applied to normalization strategies.
result Optimal normalization methods (e.g., SMC and STV) differ based on trader types. A new flow-based Bayesian filter tackles high-dimensional nonlinear stochastic systems.
problem Bayesian filtering for high-dimensional nonlinear systems is challenging due to non-Gaussian distributions and computational limitations.
method Integrates normalizing flows to construct a latent linear state-space model with efficient density estimation and sampling.
result Demonstrates superior accuracy and efficiency in numerical experiments.
Neurons predict future scalar inputs by learning top modes of lag vectors.
problem Predicting future scalar inputs with physiological delays.
method Normal Mode Decomposition to extract independently evolving modes.
result Temporal filters of neurons correspond to left eigenvectors of a generalized eigenvalue problem.
Kalman filters are routinely used for many data fusion applications including navigation, tracking, and simultaneous localization and mapping problems. However, significant time and effort is frequently required to tune various Kalman filter model parameters, e.g. process noise covariance, pre-whitening filter models f…
Local unsupervised learning outperforms end-to-end training for image classification.
problem Local unsupervised learning is believed to be inferior to end-to-end training.
method Designing a local algorithm to learn convolutional filters at scale on large image datasets using local Hebbian learning.
result Convolutional networks with patch normalization significantly outperform standard networks on image classification tasks.
Post-shifted BN prevents filter collapse in BN networks, improving model performance.
problem Filter collapse in BN networks reduces network capacity and harms model performance.
method Post-shifted BN (psBN) to prevent filter collapse by making BN parameters trainable again.
result psBN prevents filter collapse and increases model performance in various tasks.
A new filter design improves system identification accuracy.
problem Improving system identification accuracy for various system types.
method Generalized proportionate-type normalized subband adaptive filter (GPtNSAF) using least squares on subband errors with a sparsity penalty.
result GPtNSAF benefits from increasing subbands more than sparsity for quasi-sparse or dispersive systems, and both aspects are complementary for sparse systems.
We formulate the problem of neural network optimization as Bayesian filtering, where the observations are the backpropagated gradients. While neural network optimization has previously been studied using natural gradient methods which are closely related to Bayesian inference, they were unable to recover standard optim…
Latent FxLMS accelerates ANC by adapting along low-dimensional filter weights.
problem Improving active noise control with neural adaptive filters.
method Training an auto-encoder on filter coefficients, constraining weights to latent variables, and updating in latent space.
result Latent FxLMS converges in fewer steps with comparable error to standard FxLMS.
New particle filter estimates model evidence without bias.
problem Unbiased estimation of marginal likelihood for model comparison.
method Particle filter with rejection control.
result Unbiased estimation of marginal likelihood.
We examine some differential geometric approaches to finding approximate solutions to the continuous time nonlinear filtering problem. Our primary focus is a new projection method for the optimal filter infinite dimensional Stochastic Partial Differential Equation (SPDE), based on the direct L2 metric and on a family o…
Deep learning improves GW signal detection efficiency and robustness.
problem Traditional matched-filtering techniques are limited in detecting new GW signals.
method Optimized CNN models with techniques like batch normalization and dropout.
result CNN models are robust to the variation of GW waveform parameters.
We show implicit filter level sparsity manifests in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularization or weight decay. Through an extensive empirical study (Mehta et al., 2019) we hypothesize the mec…
Adaptive Heston model calibration using PCRLB and switching filters.
problem Estimating volatility in stochastic volatility models like Heston.
method Bayesian filtering (EKF, UKF, PF) with PCRLB for parameter estimation.
result Adaptive estimation of Heston model parameters improves volatility estimation.
FRN layer eliminates batch dependence in deep learning, improving performance across various tasks.
problem Batch Normalization's dependency on mini-batch elements can degrade performance for small batches.
method Filter Response Normalization (FRN) operates independently on each activation channel of each batch element.
result FRN layer outperforms BN and other alternatives in various settings for all batch sizes.
The paper uses MRFs to improve recommendation accuracy in collaborative filtering.
problem Improving recommendation accuracy in collaborative filtering.
method Modeling dependencies via Gaussian Markov Random Fields (MRFs) with auto-normal parameterization and pseudo-likelihood.
result The proposed approach achieved competitive ranking-accuracy and a 20% gain in accuracy on the largest data-set.
New findings suggest adversarial training does not flatten loss landscapes, challenging current intuition.
problem Understanding and improving generalization in deep learning.
method Loss surface visualization with filter normalization technique.
result Adversarial training does not result in flatter loss landscapes, challenging current intuition.
The redundancy is widely recognized in Convolutional Neural Networks (CNNs), which enables to remove unimportant filters from convolutional layers so as to slim the network with acceptable performance drop. Inspired by the linear and combinational properties of convolution, we seek to make some filters increasingly clo…
Dynamic pruning during training reduces deep network complexity without significant accuracy loss.
problem High memory and computational requirements of deep networks during training and inference.
method Dynamic pruning of convolutional filters during training, using L1 normalization for optimization.
result L1 normalization-based pruning yields up to 50% reduction in filters with minimal accuracy loss.
Improved product recommendations using deep learning.
problem Sparse customer purchasing data for personalized recommendations.
method Deep Collaborative Filtering (NCF) with latent variables and Bayesian Optimization.
result NCF achieved highest NDCG performance on proprietary dataset.
We study a distributionally robust mean square error estimation problem over a nonconvex Wasserstein ambiguity set containing only normal distributions. We show that the optimal estimator and the least favorable distribution form a Nash equilibrium. Despite the non-convex nature of the ambiguity set, we prove that the …
We propose two sparsity-aware normalized subband adaptive filter (NSAF) algorithms by using the gradient descent method to minimize a combination of the original NSAF cost function and the l1-norm penalty function on the filter coefficients. This l1-norm penalty exploits the sparsity of a system in the coefficients upd…
Batch normalization biases linear models towards uniform margins, improving performance in binary classification.
problem Understanding the implicit bias of batch normalization in linear models and neural networks.
method Analyzing gradient descent convergence on linear models and two-layer CNNs with batch normalization.
result Gradient descent with batch normalization in linear models converges to a uniform margin classifier with an exponential convergence rate.
CG-EnKF and NS-EnKF outperform deep learning-based SF in data assimilation.
problem Data assimilation with non-linear perturbations.
method Two non-linear extensions of EnKF: CG-EnKF and NS-EnKF.
result CG-EnKF and NS-EnKF outperform SF in high-dimensional multiscale data assimilation.
We investigate filter level sparsity that emerges in convolutional neural networks (CNNs) which employ Batch Normalization and ReLU activation, and are trained with adaptive gradient descent techniques and L2 regularization or weight decay. We conduct an extensive experimental study casting our initial findings into hy…
FPCA optimizes fairness in target vectors' span.
problem Fairness in principal component analysis for multiple target vectors.
method Non-concave maximization of worst projected target norm using sub-gradient descent.
result Optimization landscape is benign with globally optimal local minima.
Data poisoning attacks can fool neighborhood-based recommender systems.
problem Robustness of neighborhood-based recommender systems to data poisoning attacks.
method Proposed a novel data poisoning attack framework (UNAttack) to fool the systems.
result Neighborhood-based recommender systems can be easily fooled by well-designed fake users.
A filtered manifold is a smooth manifold M together with a filtration of the tangent bundle by smooth subbundles which is compatible with the Lie bracket of vector fields in a certain sense. The Lie bracket of vector fields then induces a bilinear operation on the associated graded of each tangent space of M making…
Paper uses VAEs to detect radar targets in complex noise.
problem Detecting radar targets in compound clutter and thermal noise.
method Proposes a VAE architecture to distinguish radar targets from various noise types.
result The VAE outperforms classical detectors in challenging noise conditions.
A new KF handles outliers without MSE loss.
problem Outliers degrade Kalman filter performance.
method NUV priors, EM and AM for variance estimation.
result Outlier-insensitive KF outperforms existing methods.
Surgery obstruction of a normal map to a simple Poincare pair (X,Y) lies in the relative surgery obstruction group L∗(π1(Y)→π1(X)). A well known result of Wall, the so called π-π theorem, states that in higher dimensions a normal map of a manifold with boundary to a simple Poincare pair with $π_1(X)\congπ_…
Paper analyzes robustness of data-selective Volterra NLMS algorithm.
problem Robustness analysis of data-selective Volterra NLMS algorithm.
method The paper analyzes the local robustness and proposes a global bound for the error in the coefficient vector.
result The DS-VNLMS algorithm is robust against noise and improves parameter estimation for most iterations.
FAKI improves gradient-free inference for inverse problems.
problem Expensive forward models without gradients.
method Temperature annealing with normalizing flows.
result Dramatic improvements in accuracy over EKI.
System classifies lung CT scans into normal or COVID-19 using machine learning.
problem Detecting COVID-19 infection in lung CT scans.
method MLS with CBA+KE thresholding, feature extraction, selection, and classification.
result SVM with FFV achieved 89.80% detection accuracy.
New distances for comparing multivariate normal distributions.
problem Comparing multivariate normal distributions efficiently and accurately.
method Approximated Fisher-Rao distance and pullback SPD cone distances.
result Efficient computation of distances between normal distributions.
Convolution is a central operation in Convolutional Neural Networks (CNNs), which applies a kernel to overlapping regions shifted across the image. However, because of the strong correlations in real-world image data, convolutional kernels are in effect re-learning redundant data. In this work, we show that this redund…
We developed a new approach for the analysis of physiological time series. An iterative convolution filter is used to decompose the time series into various components. Statistics of these components are extracted as features to characterize the mechanisms underlying the time series. Motivated by the studies that show …
This paper uses a geometric approach to understand how normalization layers affect neural network optimization.
problem Understanding the effect of normalization layers on optimization in neural networks.
method Introduces a spherical framework to study optimization dynamics of neural networks with normalization layers from a geometric perspective.
result Derives the first effective learning rate expression of Adam and shows that SGD with NLs is equivalent to a constrained variant of Adam.
We used convolutional neural networks (CNNs) for automatic sleep stage scoring based on single-channel electroencephalography (EEG) to learn task-specific filters for classification without using prior domain knowledge. We used an openly available dataset from 20 healthy young adults for evaluation and applied 20-fold …
Bayesian method improves neural net convergence for character recognition.
problem Improving convergence rate of neural network training algorithms.
method Customization of Kalman filter into Bayesian statistics for initialization of weights.
result Improved convergence rate for backpropagation training algorithm.
New method learns decisions from collective preferences without individual covariates.
problem Making decisions online without individual covariates.
method Collaborative filtering, matrix completion bandit, ε-greedy policy, online gradient descent, inverse propensity weighting.
result Method outperforms benchmarks and reveals new discoveries.
A minimalist approach improves LLM reasoning by filtering incorrect responses.
problem Improving large language model (LLM) reasoning on complex tasks.
method Revisit GRPO from a reinforce-like algorithm perspective, proposing Reinforce-Rej.
result RAFT, a simple rejection sampling baseline, outperforms GRPO and PPO.
The first purpose of this note is to comment on a recent article of Bursztyn, Lima and Meinrenken, in which it is proved that if M is a smooth submanifold of a manifold V, then there is a bijection between germs of tubular neighborhoods of M and germs of "Euler-like" vector fields on V. We shall explain how to approach…
A new KD model for collaborative filtering improves top-N recommendation performance.
problem Challenges in applying KD to recommender models due to feedback sparsity and ambiguity.
method Proposes a new KD model (CD) for collaborative filtering, reformulating a loss function, using probabilistic rank-aware sampling, and developing training strategies.
result Outperforms state-of-the-art methods by 2.7-33.2% in hit rate (HR) and 2.7-29.1% in NDCG.
New technique trains deep neural networks without normalization or minibatch statistics.
problem Training deep neural networks at high learning rates without normalization.
method Channel-wise zero-mean initialization and gradient modification to maintain common mode rejection.
result Achieves higher accuracy compared to batch normalization and shows minibatches are unnecessary.
Automatically designs normalization and activation layers together.
problem Designing normalization and activation layers separately.
method Unified tensor-to-tensor computation graph, low-level mathematical functions, rejection protocols, multi-objective evolution.
result Discovery of EvoNorms with novel structures.
This paper focuses on spectral filters on graphs, namely filters defined as elementwise multiplication in the frequency domain of a graph. In many graph signal processing settings, it is important to transfer a filter from one graph to another. One example is in graph convolutional neural networks (ConvNets), where the…