Combines Integrated Gradients and PatternAttribution into PGIG, outperforming alternatives.
problem Improving neural network explainability methods.
method Combines Integrated Gradients and PatternAttribution into Pattern-Guided Integrated Gradients (PGIG).
result PGIG outperforms other explainability methods in a large-scale image degradation experiment.
We demystify attention patterns in multi-head softmax models for linear data.
problem Understanding the training dynamics and emergent patterns in multi-head softmax attention models.
method Extensive empirical experiments and rigorous theoretical analysis.
result Multi-head softmax attention models approximate a debiased gradient descent predictor, outperforming single-head attention and achieving near-Bayesian optimality.
This study explains how adversarial interaction creates non-homogeneous patterns using a pseudo-Reaction-Diffusion model.
problem Understanding how adversarial interaction leads to non-homogeneous patterns in systems.
method Developed a pseudo-Reaction-Diffusion model to explain the mechanism.
result Turing instability is involved in creating non-homogeneous patterns.
A central challenge faced by memory systems is the robust retrieval of a stored pattern in the presence of interference due to other stored patterns and noise. A theoretically well-founded solution to robust retrieval is given by attractor dynamics, which iteratively clean up patterns during recall. However, incorporat…
The paper identifies patterns in language model weights used for memorizing paragraphs.
problem Locating the specific mechanisms and weights used by language models to memorize paragraphs.
method Examined gradients and attention patterns in language models to identify memorized paragraphs.
result Gradients of memorized paragraphs have a distinguishable spatial pattern, and localized attention heads are involved in paragraph memorization.
This work improves dictionary learning speed without sacrificing accuracy.
problem Prohibitive computational cost of standard dictionary learning methods.
method Approximate dictionary learning using unrolling and gradient descent.
result Unrolling outperforms standard methods in support estimation and early iterations.
Unified analysis of FL with arbitrary client participation.
problem Challenges of intermittent client availability and efficiency in FL.
method Introduces a generalized FedAvg and novel analysis capturing client participation.
result Unified convergence upper bounds for various participation patterns.
We present a novel view of nonlinear manifold learning using derivative-free optimization techniques. Specifically, we propose an extension of the classical multi-dimensional scaling (MDS) method, where instead of performing gradient descent, we sample and evaluate possible "moves" in a sphere of fixed radius for each …
Neural networks plateau during training, identified and quantified.
problem Plateau phenomenon in gradient descent training of ReLU networks.
method Identification and quantification of plateau phenomenon; new iterative training method ANLS.
result Plateaux correspond to periods of constant activation patterns; quantification of gradient flow dynamics; characterization of stationary points.
Label noise may affect the generalization of classifiers, and the effective learning of main patterns from samples with noisy labels is an important challenge. Recent studies have shown that deep neural networks tend to prioritize the learning of simple patterns over the memorization of noise patterns. This suggests a …
The study of hierarchy in networks of the human brain has been of significant interest among the researchers as numerous studies have pointed out towards a functional hierarchical organization of the human brain. This paper provides a novel method for the extraction of hierarchical connectivity components in the human …
The paper proposes a new algorithm to select subsets of training data for better accuracy and explainability.
problem Tackles the challenge of balancing accuracy and explainability in pattern recognition.
method Identifies multiple subsets with simple local patterns by clustering similar instances.
result The sub-setting algorithm outperformed traditional decision trees by 15% on the international stroke dataset.
NNNH uses neural networks to model complex event patterns.
problem Analyzing multi-dimensional nonlinear Hawkes processes with mutual excitation and inhibition.
method NNNH employs feedforward neural networks to model individual kernels and base intensity, optimizing parameters via Stochastic Gradient Descent.
result NNNH accurately captures complexities of nonlinear Hawkes processes, as demonstrated by numerical experiments.
A new associative memory uses Sinkhorn divergence for efficient pattern retrieval.
problem Efficiently retrieving patterns from large datasets of weighted point clouds.
method Derived retrieval dynamics as a SHK gradient flow, discretized for a deterministic algorithm.
result Proved basin invariance, geometric convergence, and robust recovery from perturbations.
Gradient boosting detects insider purchases predicting abnormal returns in microcap stocks.
problem Predicting abnormal returns in U.S. microcap stocks using insider purchase signals.
method Gradient boosting classifier trained on insider identity, transaction history, and market conditions.
result Gradient boosting classifier achieves AUC of 0.70 on out-of-sample data, with precision and recall of 0.38 and 0.69 respectively.
New metric m-coherence measures gradient alignment during training, revealing surprising memorization patterns.
problem Measuring and understanding the alignment of per-example gradients during training.
method Introducing m-coherence as a metric to study gradient alignment, showing its advantages over existing metrics. result Training with random labels leads to high m-coherence, indicating common patterns even when generalization is not possible. Efficiently factorizes coupled matrix tensor data for better accuracy and speed.
problem Poor computation efficiency in existing N-CMTF algorithms.
method Column-wise element selection to prevent frequent gradient updates.
result More accurate and computationally efficient factorization.
Cross-regularization adapts model complexity during training.
problem Manual tuning of model complexity for overfitting prevention.
method Directly adapts regularization parameters through validation gradients during training.
result Organic emergence of architecture-specific regularization during training.
By and large the behavior of stochastic gradient is regarded as a challenging problem, and it is often presented in the framework of statistical machine learning. This paper offers a novel view on the analysis of on-line models of learning that arises when dealing with a generalized version of stochastic gradient that …
In contrast to the standard classification paradigm where the true class is given to each training pattern, complementary-label learning only uses training patterns each equipped with a complementary label, which only specifies one of the classes that the pattern does not belong to. The goal of this paper is to derive …
This paper tackles discontinuous neural networks for better approximation of piecewise continuous functions.
problem Limitation of neural networks in approximating piecewise continuous functions due to discontinuities.
method Proposes a decoupled two-step procedure to train a discontinuous deep neural network model.
result Provides approximation guarantees for the proposed model in piecewise continuous function spaces.
Deep learning for missing data with flexible modelling.
problem Missing data with missingness dependent on the missing values.
method Deep latent variable models with a neural network for missingness pattern modeling, importance-weighted variational inference, and stochastic gradients.
result Explicitly modelling missingness improves inference on various data sets.
Neural networks learn clean data patterns first, then noisy data, leading to improved performance initially but deteriorating later.
problem Improvement in prediction error on clean data during early training of neural networks with noisy labels.
method Theoretical analysis and experiments to explore the dynamics of gradient descent and the impact of clean and noisy data.
result Neural networks prioritize learning clean data patterns first, leading to improved performance initially but deteriorating later due to diminishing gradient dominance of clean samples over noisy ones.
Classification and regression in which the inputs are graphs of arbitrary size and shape have been paid attention in various fields such as computational chemistry and bioinformatics. Subgraph indicators are often used as the most fundamental features, but the number of possible subgraph patterns are intractably large …
Introduces new gradient-based methods for machine learning problems.
problem New challenges in machine learning due to decision-making and multi-agent problems.
method Gradient-based optimization and variational inequalities.
result Shifts focus from pattern recognition to decision-making and multi-agent problems.
Gradient-based methods improve understanding of deep learning survival models.
problem Limited interpretability of deep learning survival models hinders their adoption.
method Gradient-based explanation methods tailored to survival neural networks.
result Gradient-based methods capture feature effects and temporal dynamics.
Time series motifs play an important role in the time series analysis. The motif-based time series clustering is used for the discovery of higher-order patterns or structures in time series data. Inspired by the convolutional neural network (CNN) classifier based on the image representations of time series, motif diffe…
XGBoost detects unlawful insider trading with high accuracy.
problem Detecting unlawful insider trading from large volumes of transactions.
method Applying eXtreme Gradient Boosting (XGBoost) for identifying and ranking key features.
result XGBoost achieves 97% accuracy in detecting unlawful transactions.
In this work we propose Hebbian-descent as a biologically plausible learning rule for hetero-associative as well as auto-associative learning in single layer artificial neural networks. It can be used as a replacement for gradient descent as well as Hebbian learning, in particular in online learning, as it inherits the…
New method optimizes MRI sampling patterns for faster scans.
problem Accelerate MRI scans without sacrificing image quality.
method Joint learning of adaptive sampling patterns and model-based recovery.
result Improved MR image quality compared to other methods.
Analytical method finds deeper optima in two-layer ReLU networks.
problem Training two-layer ReLU networks with analytical methods.
method Analytically finding critical points of the loss function for one layer while keeping the other fixed.
result Significantly smaller training loss values on real datasets compared to gradient descent methods.
This paper proposes Evolutionary Multi-objective Optimization (EMO)-based Adversarial Example (AE) design method that performs under black-box setting. Previous gradient-based methods produce AEs by changing all pixels of a target image, while previous EC-based method changes small number of pixels to produce AEs. Than…
PEGR improves deep learning models' robustness against noisy data.
problem Learning signals from noisy data in deep learning models.
method Per-example gradient regularization (PEGR) to suppress noise.
result PEGR enhances test error and robustness against noise perturbations.
A new algorithm estimates sparse gradients on graphs with improved risk bounds.
problem Estimating sparse gradients on graph-structured data.
method Tree-Projected Gradient Descent algorithm for gradient-sparse parameters.
result Achieves risk bound of ns∗log(1+s∗p). We study the implicit bias of gradient descent methods in solving a binary classification problem over a linearly separable dataset. The classifier is described by a nonlinear ReLU model and the objective function adopts the exponential loss function. We first characterize the landscape of the loss function and show th…
Recurrent Neural Networks (RNN), Long Short-Term Memory Networks (LSTM), and Memory Networks which contain memory are popularly used to learn patterns in sequential data. Sequential data has long sequences that hold relationships. RNN can handle long sequences but suffers from the vanishing and exploding gradient probl…
Paper uses Super-App data to improve income estimation models.
problem Improving accuracy of income estimation models.
method TreeSHAP method for Stochastic Gradient Boosting Interpretation.
result Alternative data from Super-Apps capture more information than traditional financial data.
BOOOM optimizes orthonormal matrices without needing gradients.
problem Optimizing over the Stiefel manifold in non-convex, non-smooth settings.
method Global Givens rotation-based parametrization and Recursive Modified Pattern Search.
result BOOOM achieves strong performance across various optimization problems.
We consider a class of a nested optimization problems involving inner and outer objectives. We observe that by taking into explicit account the optimization dynamics for the inner objective it is possible to derive a general framework that unifies gradient-based hyperparameter optimization and meta-learning (or learnin…
Novel PCA method for high-dimensional inverse problems.
problem Optimizing large-scale random fields with gradient information.
method Gradient-Sensitive Principal Component Analysis (Gradient-SPCA) that modifies PCA using objective function gradients.
result Improvements in encoding quality for objective function minimization and field distribution.
Learning rate decay (lrDecay) is a \emph{de facto} technique for training modern neural networks. It starts with a large learning rate and then decays it multiple times. It is empirically observed to help both optimization and generalization. Common beliefs in how lrDecay works come from the optimization analysis of (S…
Express Wavenet is an improved optical diffractive neural network. At each layer, it uses wavelet-like pattern to modulate the phase of optical waves. For input image with n2 pixels, express wavenet reduce parameter number from O(n2) to O(n). Only need one percent of the parameters, and the accuracy is still very high.…
Global models outperform univariate benchmarks in complex time series forecasting.
problem Comparing global forecasting models to univariate benchmarks in various challenging scenarios.
method Simulated datasets with controlled characteristics, including homogeneity, complexity, and series lengths. Global forecasting models (RNN, LGBM) compared to univariate techniques.
result Global models like RNN and LGBM are competitive in complex scenarios with short series lengths and heterogeneous data.
This paper uses deep reinforcement learning to optimize stock portfolios considering transaction costs and risks.
problem Optimizing stock portfolios with transaction costs and risks.
method Formulated stock portfolio optimization as a reinforcement learning problem, applied DDPG, GDPG, and PPO algorithms, and used Wavelet Transform.
result DDPG and GDPG algorithms outperformed PPO in continuous action space.
Paper proposes SCott optimizer to reduce forecasting model training variance.
problem Large variance in gradient estimation for forecasting models.
method Stratified sampling and control variate to reduce gradient variance.
result SCott optimizer converges faster on time series forecasting problems.
We examine the role of memorization in deep learning, drawing connections to capacity, generalization, and adversarial robustness. While deep networks are capable of memorizing noise data, our results suggest that they tend to prioritize learning simple patterns first. In our experiments, we expose qualitative differen…
The typical multi-task learning methods for spatio-temporal data prediction involve low-rank tensor computation. However, such a method have relatively weak performance when the task number is small, and we cannot integrate it into non-linear models. In this paper, we propose a two-step suboptimal unitary method (SUM) …
Adam performs better with equal momentum parameters, revealing a gradient scale invariance principle.
problem Why Adam performs better with β1=β2. method Formalized gradient scale invariance and proved it for Adam with equal β1 and β2. result Adam becomes gradient scale invariant of first order if and only if β1=β2.