Gradient descent with proportional updates addresses training instabilities in deep learning.
problem Training instabilities like vanishing and exploding gradients in deep learning.
method Gradient descent with updates proportional to weight norms, independent of gradient magnitudes.
result Theoretical analysis confirms the effectiveness of gradient descent with proportional updates.
EGMU optimizes portfolios using KL divergence, ensuring positive solutions.
problem Constructing multi-factor target-exposure portfolios efficiently and accurately.
method Convex optimization framework minimizing KL divergence, with explicit solvers.
result Established feasibility and uniqueness of strictly positive solutions under convex-hull conditions.
Humans are able to accelerate their learning by selecting training materials that are the most informative and at the appropriate level of difficulty. We propose a framework for distributing deep learning in which one set of workers search for the most informative examples in parallel while a single worker updates the …
Develops efficient method for updating models with small data changes.
problem Efficiently updating models when data changes (e.g., adding/removing instances/features).
method Generalized Low-Rank Update (GLRU) for non-linear estimators.
result Provides updated solutions with computational complexity proportional to dataset changes.
We describe k-MLE, a fast and efficient local search algorithm for learning finite statistical mixtures of exponential families such as Gaussian mixture models. Mixture models are traditionally learned using the expectation-maximization (EM) soft clustering technique that monotonically increases the incomplete (expec…
Aioli unifies language model data mixing methods and improves performance.
problem Optimizing the mixture of training data groups for language models.
method Unified optimization framework for dynamically adjusting mixture proportions.
result Aioli outperforms existing methods by up to 12.012 test perplexity points.
We argue that the existing regret matchings for Nash equilibrium approximation conduct "jumpy" strategy updating when the probabilities of future plays are set to be proportional to positive regret measures. We propose a geometrical regret matching which features "smooth" strategy updating. Our approach is simple, intu…
Optimal best-arm identification in linear bandits reduces sampling budget.
problem Identifying the best arm with fixed confidence in stochastic linear bandits.
method A simple algorithm that tracks an optimal proportion of arm draws, updated as rarely as desired.
result The algorithm's sampling complexity matches known lower bounds, asymptotically almost surely and in expectation.
When training large machine learning models with many variables or parameters, a single machine is often inadequate since the model may be too large to fit in memory, while training can take a long time even with stochastic updates. A natural recourse is to turn to distributed cluster computing, in order to harness add…
Stochastic gradient descent procedures have gained popularity for parameter estimation from large data sets. However, their statistical properties are not well understood, in theory. And in practice, avoiding numerical instability requires careful tuning of key parameters. Here, we introduce implicit stochastic gradien…
The paper improves Gibbs sampling for large graphs by minibatching.
problem High computational cost of single Gibbs sampling update step.
method Minibatching: subsampling factors to estimate their sum.
result Minibatched Gibbs can be made unbiased and converge faster.
A new method improves Bayesian inference for multimodal posteriors.
problem Insensitivity to well-separated modes in multimodal posteriors.
method Weighted Kernel Stein Discrepancy method.
result Significantly improved mode sensitivity compared to standard KSD-Bayes.
Paper introduces a simpler gated RNN structure to better capture long-term dependencies.
problem Difficulty in learning long-term dependencies in RNNs.
method Proposes a grouped distributor unit (GDU) with partitioned hidden states and adaptive update rates.
result GDU outperforms LSTM and GRU on various tasks, including pathological and natural data.
Training large machine learning (ML) models with many variables or parameters can take a long time if one employs sequential procedures even with stochastic updates. A natural solution is to turn to distributed computing on a cluster; however, naive, unstructured parallelization of ML algorithms does not usually lead t…
In recent years, dynamically growing data and incrementally growing number of classes pose new challenges to large-scale data classification research. Most traditional methods struggle to balance the precision and computational burden when data and its number of classes increased. However, some methods are with weak pr…
GD converges in unstable regimes, even with oscillatory behavior.
problem Understanding convergence of GD in unstable regimes.
method Analysis of two-step gradient updates.
result Characterization of local conditions for convergence.
RLCFR improves CFR's generalization in imperfect information games.
problem Improving CFR's performance in large-scale, imperfect information games.
method RLCFR integrates CFR with deep reinforcement learning to update strategies dynamically.
result RLCFR significantly enhances CFR's generalization ability in various games.
Bayesian framework improves minority class performance in class-imbalanced data.
problem Class imbalance in predictive toxicology models.
method Weighted likelihood approach modifying likelihood function weights inversely proportional to class proportions.
result Improves balanced accuracy and sensitivity for minority class (toxic compounds).
A method for selecting pseudo-labeled data in semi-supervised learning using generalized Bayes and soft revision.
problem Selecting pseudo-labeled data for semi-supervised learning with robustness to uncertainty.
method Using credal sets and the Gamma-Maximin method with soft revision to update priors and select pseudo-labeled data.
result The Gamma-Maximin method with soft revision can achieve promising results, especially in scenarios with low labeled data proportions.
In this paper we develop an asymptotic analysis for formal and actual solutions of q-difference equations, under a regularity assumption. In particular, evaluations of regular solutions of regular q-difference equations have an exponential growth rate which can be computed from the q-difference equation. The motivation…
Paper proposes a method to estimate true positive proportion without knowing it.
problem Bias in binary classifier performance due to different positive item proportions.
method Maximum likelihood estimator for true proportion of positives.
result Method accurately estimates true positive proportion in data sets.
We consider object recognition in the context of lifelong learning, where a robotic agent learns to discriminate between a growing number of object classes as it accumulates experience about the environment. We propose an incremental variant of the Regularized Least Squares for Classification (RLSC) algorithm, and expl…
Combines proportional and stop-loss reinsurance for insurer and reinsurer.
problem Addressing conflicting interests between insurer and reinsurer.
method Introduces proportional-stop-loss reinsurance using balanced loss function.
result Maximizes expected surplus for both insurer and reinsurer.
We investigate how and when to diversify capital over assets, i.e., the portfolio selection problem, from a signal processing perspective. To this end, we first construct portfolios that achieve the optimal expected growth in i.i.d. discrete-time two-asset markets under proportional transaction costs. We then extend ou…
Information-Geometric Optimization (IGO) is a unified framework of stochastic algorithms for optimization problems. Given a family of probability distributions, IGO turns the original optimization problem into a new maximization problem on the parameter space of the probability distributions. IGO updates the parameter …
Improved Lasso estimator speeds up variable selection.
problem Efficient variable selection in high-dimensional data.
method Stability principle-based generalized debiased Lasso.
result Significantly reduces computational cost of resampling-based methods.
Paper improves deep learning for instance-level classification from label proportions.
problem Dealing with noisy pseudo-labeling and high-entropy class distributions in LLP.
method Introducing a two-stage training approach with constrained optimization and mixup strategy.
result Significant performance improvement in instance-level classification.
SQUEAK reduces space complexity for Nystrom approximations in KRR.
problem Large datasets in KRR require impractical storage space.
method SQUEAK uses unnormalized ridge leverage scores for incremental updates.
result Space complexity improved with constant factor worse than exact RLS.
MSTGD optimizes gradient descent with stratified sampling for faster convergence.
problem Fluctuation in gradient expectation and variance between iterations.
method Memory Stochastic Stratified Gradient Descent (MSTGD) with stratified sampling and variance reduction.
result MSTGD achieves an exponential convergence rate independent of dataset size and batch size.
CMCD sampler connects transport and variational inference for efficient sampling.
problem Efficient sampling and generative modeling in Bayesian computation.
method Developed a principled framework using divergences on path space, CMCD sampler with adaptive dynamics.
result CMCD sampler outperforms competing approaches across various experiments.
Learning from Label Proportions (LLP) is a learning setting, where the training data is provided in groups, or "bags", and only the proportion of each class in each bag is known. The task is to learn a model to predict the class labels of the individual instances. LLP has broad applications in political science, market…
Develops a method to estimate average hazard under non-proportional hazards without relying on proportional hazards assumption.
problem Estimation of treatment effects when hazards are non-proportional, leading to unstable hazard ratios.
method Semiparametric, doubly robust framework for covariate-adjusted average hazard estimation.
result Valid sqrt{n} inference with small bias and near-nominal confidence-interval coverage across proportional and non-proportional hazards settings.
MetaGrad adapts multiple learning rates for faster online optimization.
problem Online convex optimization with varying function types.
method MetaGrad uses multiple adaptive learning rates based on empirical performance.
result MetaGrad achieves faster rates for various function types.
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.
New algorithm samples matrix rows proportional to their ℓ_p norm in a turnstile data stream.
problem Sampling rows of a dynamic matrix efficiently in a turnstile data stream.
method Develops a novel algorithm for sampling rows proportional to their ℓ_p norm in a turnstile data stream, returning sampled row indexes and approximated sampling probabilities.
result Achieves (1+ε) approximation for logistic regression in a turnstile data stream with polynomial sketch size. A model learns set comparison, vague quantification, and proportional estimation from visual scenes.
problem Learning of quantification mechanisms from visual scenes.
method A multi-task computational model that integrates set comparison, vague quantification, and proportional estimation.
result The multi-task model performs better when lower-complexity tasks are available and can generalize to unseen combinations.
New learning rules achieve optimal sample complexity for weakly supervised classification.
problem Learning from label proportions in weakly supervised settings.
method Debiased proportional square loss and EasyLLP learning rule.
result Achieves optimal sample complexity in both realizable and agnostic settings.
DiffusionBlocks trains neural networks by breaking them into independent blocks, reducing memory usage.
problem Memory bottlenecks in end-to-end neural network training.
method Transforming transformer-based networks into independent trainable blocks via a denoising process.
result Independent block-wise training matches end-to-end training performance while reducing memory requirements.
LDTA expands LDA's topic modeling capacity with tree-structured priors.
problem Limited expressiveness of Dirichlet priors in LDA for complex topic relationships.
method Introduces Latent Dirichlet-Tree Allocation (LDTA) with Dirichlet-Tree (DT) priors, and develops universal mean-field variational inference and Expectation Propagation.
result LDTA enables expressive, tree-structured priors over topic proportions, expanding modeling capacity of LDA.
learn2mix trains neural nets faster by adjusting class proportions dynamically.
problem Training neural nets efficiently with limited resources and imbalanced classes.
method Adaptive class proportion adjustment during training.
result Neural nets trained with learn2mix converge faster than static methods.
Algorithm learns from label proportions in unlabeled bags.
problem Learning from unlabeled bags with known label proportions.
method Differentiable loss functions for deep neural networks.
result Deep neural networks can accurately classify images from unlabeled bags.
A flexible method for estimating mixture proportions in PU learning.
problem Estimating the proportion of positive samples in PU learning.
method Construct a probabilistic classifier and apply a one-dimensional mixture proportion method.
result Consistent estimators with competitive performance on simulated and real data.
We study the problem of learning with label proportions in which the training data is provided in groups and only the proportion of each class in each group is known. We propose a new method called proportion-SVM, or ∝SVM, which explicitly models the latent unknown instance labels together with the known group …
Study examines how insurance affects households prone to proportional losses, especially those near poverty.
problem Impact of insurance on households susceptible to proportional losses, focusing on poverty traps.
method Modelled proportional capital losses with insurance, derived closed formulae and non-local differential equations.
result New formulae and methods to calculate trapping probability, constraints on parameters to prevent certainty of trapping.
Paper proposes a transfer learning method for learning with label proportions that handles uncertain data.
problem Learning with label proportions (LLP) for uncertain data.
method Transfer learning approach to transfer knowledge from source to target tasks with uncertain data.
result The proposed TL-LLP method achieves better accuracy and is less sensitive to noise.
Paper proposes a new method for estimating mixture proportions without irreducibility assumption.
problem Estimating mixture proportions when component distributions are not irreducible.
method Developed a resampling-based meta-algorithm that adapts existing MPE algorithms to non-irreducible settings.
result Empirical results show improved estimation performance compared to baseline methods and regrouping-based algorithms.
Generative adversarial networks learn from label proportions without distributional restrictions.
problem Learning from label proportions with limited information.
method Generative adversarial networks (GANs) for LLP-GAN, approximating label proportions without distributional assumptions.
result Global optimality of LLP-GAN proved under mild assumptions.
Proportional centroid clustering aims to fairly group points without prior protected subsets.
problem Fairly group points without prior protected subsets.
method Define fairness as proportionality, present algorithms for efficient computation and optimization.
result Proportional solutions trade off with the k-means objective.