Continuous-time SGD converges under certain conditions, useful for deep learning.
problem Minimizing population expected loss in learning problems.
method Continuous-time approximation of stochastic gradient descent.
result Establishes sufficient conditions for convergence, applicable to overparametrized neural networks.
Paper analyzes agnostic learning of mixed linear regression without generative models.
problem Learning mixed linear regression without assuming stochastic generation.
method Expectation Maximization (EM) and Alternating Minimization (AM) algorithms.
result AM and EM algorithms converge to population loss minimizers under standard conditions.
Training activation quantized neural networks involves minimizing a piecewise constant function whose gradient vanishes almost everywhere, which is undesirable for the standard back-propagation or chain rule. An empirical way around this issue is to use a straight-through estimator (STE) (Bengio et al., 2013) in the ba…
Study asset pricing with reference-dependent preferences, finding matching equity premia.
problem Understanding asset pricing under reference-dependent preferences.
method Discrete-time consumption-based capital asset pricing model with reference-dependent preferences.
result Models can generate equity premia matching empirical estimates, showing procyclical price-dividend ratio and countercyclical equity premium.
Develops certificates for local population-risk increments using cross-fitted ridge calibration.
problem Certifying local population-risk increments in statistical models.
method Cross-fitted ridge calibration for linear feature classes, separating Taylor fluctuations and remainders.
result Certifies measurable updates from the same sample with penalties dependent on empirical geometry.
We study differentially private (DP) algorithms for stochastic convex optimization (SCO). In this problem the goal is to approximately minimize the population loss given i.i.d. samples from a distribution over convex and Lipschitz loss functions. A long line of existing work on private convex optimization focuses on th…
Paper addresses DP-SCO on heavy-tailed data, providing methods and results.
problem Designing DP algorithms for SCO on heavy-tailed data.
method Sample-and-aggregate framework, gradient smoothing and trimming.
result Achieved DP guarantees for various loss functions with different excess population risks.
Though machine learning algorithms excel at minimizing the average loss over a population, this might lead to large discrepancies between the losses across groups within the population. To capture this inequality, we introduce and study a notion we call maximum weighted loss discrepancy (MWLD), the maximum (weighted) d…
New framework controls statistical dispersion for high-stakes applications.
problem Understanding and controlling the dispersion of loss distributions in high-stakes applications.
method Simple yet flexible framework for distribution-free control of statistical dispersion measures.
result Proposed methods control statistical dispersion measures with societal implications.
Uncertainty sampling, a popular active learning algorithm, is used to reduce the amount of data required to learn a classifier, but it has been observed in practice to converge to different parameters depending on the initialization and sometimes to even better parameters than standard training on all the data. In this…
A clustering method for multivariate populations with similar dependence structures.
problem Grouping populations with similar dependence structures.
method Orthogonal projection coefficients of density copulas estimated from populations.
result Clusters of populations with similar dependence structures.
Proposes φ-balancing for more balanced expert utilization in MoE models.
problem Balanced expert utilization in MoE models to avoid bias.
method Directly targets population-level balance by minimizing a convex potential function.
result Consistently outperforms prior methods in stability and effectiveness.
Given samples from a distribution, how many new elements should we expect to find if we continue sampling this distribution? This is an important and actively studied problem, with many applications ranging from unseen species estimation to genomics. We generalize this extrapolation and related unseen estimation proble…
The paper tackles optimal policy learning with asymmetric counterfactual utilities in healthcare decisions.
problem Learning optimal policies from observed data with asymmetric counterfactual utilities.
method The approach involves identifying and minimizing the maximum expected utility loss using statistical decision theory and solving intermediate classification problems.
result One can learn minimax loss decision rules from observed data.
Minimizing a convex risk function is the main step in many basic learning algorithms. We study protocols for convex optimization which provably leak very little about the individual data points that constitute the loss function. Specifically, we consider differentially private algorithms that operate in the local model…
Polyak step size GD reaches final radius of convergence after log iterations.
problem Statistical and computational complexities of Polyak step size GD.
method Generalized smoothness and Lojasiewicz conditions, stability of gradients.
result Polyak step size GD reaches final statistical radius of convergence after logarithmic number of iterations.
Stochastic Gradient Descent can overfit after just a few passes, contrary to initial expectations.
problem Understanding the out-of-sample performance of multi-pass SGD in stochastic convex optimization.
method Analysis of multi-pass SGD in the stochastic convex optimization model.
result Multi-pass SGD can lead to significant overfitting after just a few passes, contrary to initial expectations.
Improved similarity search in embeddings using InfoNCE loss.
problem Improving similarity search in embedding models trained by contrastive learning.
method Introduced a new continuity bound for InfoNCE loss via Gâteaux differentiation, preserving the averaging effect of negative samples.
result Demonstrated that the averaging effect of k negative samples in InfoNCE loss carries over to stabilisation of generalisation error as k grows. Gradient EM converges exponentially to optimal solution in agnostic mixtures.
problem Fitting k parametric functions to given data points without a generative model. method Gradient EM algorithm for agnostic mixtures of arbitrary parametric functions.
result Gradient EM converges exponentially to population loss minimizers with high probability.
Paper tackles estimating individual treatment effects from observational data.
problem Estimating the difference between outcomes with and without treatment from single observation.
method Formulated as inference from hidden variables, uses a model of four causal populations, proposes ECM algorithm.
result ECM algorithm provides better performance compared to baseline methods on synthetic and real-world data.
New DP algorithm improves privacy and efficiency for convex optimization.
problem Efficient, DP algorithms for convex optimization with strong excess risk bounds.
method Output perturbation for a broad class of tilted loss functions.
result Near optimal DP excess risk and runtime bounds for convex optimization.
A new method identifies sub-populations in unlabelled heterogeneous data by accounting for co-features.
problem Estimating sub-populations in unlabelled heterogeneous data with co-features.
method Mixture of Conditional Gaussian Graphical Models (CGGM) with penalized EM algorithm.
result The method successfully identifies sub-populations disrupted by co-features.
The study investigates the consistency of k-means clustering under finite expectation assumptions.
problem Consistency of k-means clustering under finite expectation assumptions. method Investigates the conditions under which k-means clustering is consistent, considering finite expectation instead of finite variance. result Inconsistency can arise due to extreme cluster imbalance, leading to some clusters having few points.
In this paper, a new population-guided parallel learning scheme is proposed to enhance the performance of off-policy reinforcement learning (RL). In the proposed scheme, multiple identical learners with their own value-functions and policies share a common experience replay buffer, and search a good policy in collabora…
We consider the \mnk{classical} problem of a controller activating (or sampling) sequentially from a finite number of N≥2 populations, specified by unknown distributions. Over some time horizon, at each time n=1,2,…, the controller wishes to select a population to sample, with the goal of sampling fro…
We introduce a deep neural network based method for solving a class of elliptic partial differential equations. We approximate the solution of the PDE with a deep neural network which is trained under the guidance of a probabilistic representation of the PDE in the spirit of the Feynman-Kac formula. The solution is giv…
Study merges datasets to improve AI model performance.
problem Improving machine learning models with diverse datasets.
method Developed an algorithm using oracle inequality and data-driven estimators.
result Algorithm reduces population loss with high probability.
Gradient descent learns over-param neural nets better than NTK.
problem Learning over-parametrized neural networks with ReLU activations.
method Gradient descent from random initialization on a Gaussian input distribution.
result Gradient descent achieves population loss o(1/d), while NTK achieves Ω(1/d). Gradient descent learns a single neuron without knowing the relationship between inputs and labels.
problem Learning a single neuron without knowing the relationship between inputs and labels.
method Using gradient descent to minimize empirical risk over i.i.d. samples, with a nonconvex and nonsmooth optimization problem.
result Gradient descent achieves near-optimal population risk in polynomial time and sample complexity.
Semi-supervised EM improves convergence rate with labeled samples.
problem Improving convergence rate in EM algorithm with labeled and unlabeled data.
method Analysis of semi-supervised EM algorithm for Gaussian mixture models.
result Labeled samples significantly improve the convergence rate for the EM algorithm.
Theory integrates loss aversion into expected utility for monetary returns.
problem Modeling loss aversion in expected utility theory.
method Develops state-dependent linear utility functions incorporating loss aversion.
result Contracts from monopolists in insurance markets.
Optimum in Convex Hulls (OCH) generalizes clinical trial results to broader populations.
problem Clinical trials exclude confounding but limit recruitment; observational data are more inclusive but suffer from confounding.
method OCH uses convex hulls of conditional expectations or densities to approximate the true treatment effect from both observational and trial data.
result OCH estimates the treatment effect with state-of-the-art accuracy in terms of both expectations and densities.
A new method approximates expected empirical loss for stochastic deep learning tasks.
problem Determining optimal step sizes for stochastic gradient descent in deep learning.
method Applying one-dimensional function fitting to noisy losses of vertical cross sections to approximate expected empirical loss.
result The method leads to a robust and straightforward optimization method that performs well across datasets and architectures.
Loss-calibrated EP improves Bayesian decision-making by focusing on utility-sensitive posterior approximations.
problem Bayesian decision-making under asymmetric utility functions.
method Loss-calibrated expectation propagation (Loss-EP) that tilts the posterior towards higher utility decisions.
result Loss-EP can capture useful information for decision-making under asymmetric penalties.
Although kernel methods are widely used in many learning problems, they have poor scalability to large datasets. To address this problem, sketching and stochastic gradient methods are the most commonly used techniques to derive efficient large-scale learning algorithms. In this study, we consider solving a binary class…
The landscape of empirical risk has been widely studied in a series of machine learning problems, including low-rank matrix factorization, matrix sensing, matrix completion, and phase retrieval. In this work, we focus on the situation where the corresponding population risk is a degenerate non-convex loss function, nam…
Paper revisits DP-SCO in Euclidean and ℓpd spaces, focusing on constrained and bounded sets.
problem Differentially private stochastic convex optimization in constrained and bounded sets in Euclidean and ℓpd spaces. method Proposes methods achieving excess population risks dependent on Gaussian width of the constraint set, and novel algorithms for unconstrained and heavy-tailed data.
result Theoretical results for DP-SCO in ℓpd spaces, including optimal bounds for strongly convex functions. Investigates conditions for risk or utility functionals to be sensitive to large losses.
problem Conditions for risk or utility functionals to be sensitive to large losses.
method Analyzes sensitivity to large losses for various risk and utility functionals.
result Value at Risk and Expected Shortfall generally fail to be sensitive to large losses, but expected utility functionals and certain adjusted versions are sensitive.
New algorithms for differentially private optimization in convex and non-convex settings with near-optimal rates.
problem Differentially private optimization in convex and non-convex settings.
method Developed algorithms for convex and non-convex settings with near-optimal excess population risk.
result Achieved near-optimal rates in near-linear time for convex settings and nearly dimension independent rates for non-convex settings.
The paper analyzes the maximum margin algorithm's performance on noisy data.
problem Analyzing the performance of maximum margin algorithm on noisy data.
method Finite-sample analysis of maximum margin algorithm applied to noisy data.
result The maximum margin algorithm can achieve nearly optimal population risk with sufficient over-parameterization.
The classical asymptotic theory for parametric M-estimators guarantees that, in the limit of infinite sample size, the excess risk has a chi-square type distribution, even in the misspecified case. We demonstrate how self-concordance of the loss allows to characterize the critical sample size sufficient to guarantee …
The study optimizes bounds for comparing training and population loss.
problem Optimizing bounds for comparing training and population loss.
method Derives generic information-theoretic and PAC-Bayesian generalization bounds using convex comparator functions.
result The tightest possible bound is obtained with the comparator being the convex conjugate of the CGF of the bounding distribution.
MRCs minimize worst-case expected 0-1 loss and provide performance guarantees.
problem Minimizing expected 0-1 loss in classification.
method Minimizes worst-case expected 0-1 loss over uncertainty sets defined by linear constraints.
result Achieves efficient learning and generalization with performance guarantees.
Consider the problem of finding a population or a probability distribution amongst many with the largest mean when these means are unknown but population samples can be simulated or otherwise generated. Typically, by selecting largest sample mean population, it can be shown that false selection probability decays at an…
This paper rethinks confidence calibration under covariate shifts.
problem Calibration methods struggle with covariate shifts and unstable importance weighting.
method Derives Expectation consistency condition and proposes Expectation consistency loss (ECL).
result ECL loss is compatible with various types of calibration and has the same sample complexity as ECE.
Paper improves privacy-preserving optimization rates for convex functions.
problem Differentially private stochastic convex optimization.
method Algorithmic improvements for convex and strongly convex functions under TNC and non-negative loss.
result Excess population risk bounds for DP-SCO are faster than previous results.
Dropout training is shown to be optimal for adversarial covariate corruption.
problem Adversarial corruption of covariates in machine learning models.
method Minimax game analysis with dropout noise as the adversarial strategy.
result Dropout training minimizes expected loss for multiplicative perturbations of covariates.
The paper examines bounds for stop-loss payoffs using transformed random variables.
problem Bounding stop-loss payoffs for a difference of two random variables.
method Analyzes crossing points of cdfs of original and transformed random variables.
result Unique pairwise crossing points for mortality-linked securities under symmetric copulas.