Agghoo averages hold-out learning rules for theoretical and practical risk minimization.
problem Theoretical and practical risk minimization for various learning problems.
method Averaging hold-out learning rules.
result Agghoo performs at worst like hold-out for convex risks and classification with 0-1 risk.
This paper improves neural network predictions with early stopping using conformal calibration.
problem Lack of precise statistical guarantees for neural networks trained with early stopping.
method Conformalized early stopping that combines early stopping with conformal calibration.
result Models provide both accuracy and precise inferences without additional data splits.
One significant challenge to scaling entity resolution algorithms to massive datasets is understanding how performance changes after moving beyond the realm of small, manually labeled reference datasets. Unlike traditional machine learning tasks, when an entity resolution algorithm performs well on small hold-out datas…
A new strategy selects k in k-NN regression without hold-out data.
problem Choosing optimal k in k-NN regression without hold-out data.
method Iterative procedure over k, minimum discrepancy principle.
result Minimax-optimal over smoothness function classes.
New method improves deep learning models' uncertainty estimates.
problem Overconfidence in deep learning predictions.
method Develops a novel training algorithm using conformal inference.
result Produces more reliable uncertainty estimates without sacrificing accuracy.
The strategy of early stopping is a regularization technique based on choosing a stopping time for an iterative algorithm. Focusing on non-parametric regression in a reproducing kernel Hilbert space, we analyze the early stopping strategy for a form of gradient-descent applied to the least-squares loss function. We pro…
Improves risk control in predictions using semi-supervised calibration.
problem Noisy hyper-parameter tuning from limited labeled data.
method Semi-supervised calibration using unlabeled data to tune hyper-parameters rigorously.
result Improves prediction accuracy without sacrificing statistical validity.
The paper validates a method for recovering over-parameterized matrices and images from noisy measurements.
problem Recovering a low-rank matrix from noisy measurements when the rank is unknown.
method Using gradient descent with small random initialization on a nonconvex objective function built from a rank-overspecified factored representation of the matrix variable.
result Gradient descent iterations converge to the ground-truth matrix under certain conditions and can be stopped efficiently to detect a nearly optimal estimator.
This paper is a comment on the survey paper by Biau and Scornet (2016) about random forests. We focus on the problem of quantifying the impact of each ingredient of random forests on their performance. We show that such a quantification is possible for a simple pure forest , leading to conclusions that could apply more…
Analyzing a comprehensive news dataset, we document that joint news coverage triggers attention contagion, causing temporarily inflated valuations for affected stocks. Tracing SEC EDGAR visits from unique IPs, we provide direct evidence of attention spillovers between stocks. Stocks with greater joint news coverage exh…
New algorithm reduces regret in contextual bandits with many near-boundary contexts.
problem High regret in contextual bandits with many near-boundary contexts.
method Hierarchical nearest neighbour approach, holding out contexts for computation.
result Eradicates high regret in adversarial contextual bandits.
Bias correction needed after deep learning regression training.
problem Systematic error accumulation in deep learning regression models.
method Adjust bias of the machine learning model post-training.
result Bias correction efficiently solves error accumulation.
In this article, we derive concentration inequalities for the cross-validation estimate of the generalization error for subagged estimators, both for classification and regressor. General loss functions and class of predictors with both finite and infinite VC-dimension are considered. We slightly generalize the formali…
New method for estimating covariance with robustness to outliers.
problem Estimating covariance from noisy data with outliers.
method Cross-fitted clipped covariance estimator with computable Bernstein certificates.
result The method balances certified stochastic error and robust hold-out proxy for clipping bias.
Proposes new method for calibrating treatment effect predictors.
problem Calibrating predictors of heterogeneous treatment effects.
method Causal isotonic calibration and cross-calibration.
result Achieves fast calibration rates under weak conditions.
The paper optimizes hyperplanes for binary classification in high-dimensional data with latent Gaussian mixtures.
problem Binary classification in high-dimensional data with latent Gaussian mixtures.
method Generalized least squares estimator for estimating the direction of the optimal separating hyperplane. Simple correction for intercept estimation.
result The procedure is minimax optimal in many scenarios and can retain the interpolation property.
Adaptive classification methods ensure correct prediction intervals.
problem Developing methods to ensure correct prediction intervals for classification problems.
method Specialized conformal inference techniques combining cross-validation+, jackknife+, and a novel conformity score.
result The methods provide guaranteed approximate conditional coverage for complex data distributions.
In bankruptcy prediction, the proportion of events is very low, which is often oversampled to eliminate this bias. In this paper, we study the influence of the event rate on discrimination abilities of bankruptcy prediction models. First the statistical association and significance of public records and firmographics i…
The concept of SCN offers a fast framework with universal approximation guarantee for lifelong learning of non-stationary data streams. Its adaptive scope selection property enables for proper random generation of hidden unit parameters advancing conventional randomized approaches constrained with a fixed scope of rand…
We address the problem of {\it adaptivity} in the framework of reproducing kernel Hilbert space (RKHS) regression. More precisely, we analyze estimators arising from a linear regularization scheme $g_\lam$. In practical applications, an important task is to choose the regularization parameter $\lam$ appropriately, i.e.…
The article presents methods to select models from behavioral learning data, with applications to contextual bandits.
problem Model selection for behavioral learning data, especially in non-stationary environments.
method Two model selection methods: a general hold-out procedure and an AIC-type criterion, adapted for non-stationary dependent data.
result Theoretical error bounds for these methods are close to those of the standard i.i.d. case.
In de novo drug design, computational strategies are used to generate novel molecules with good affinity to the desired biological target. In this work, we show that recurrent neural networks can be trained as generative models for molecular structures, similar to statistical language models in natural language process…
In this paper we establish a one-to-one correspondence between S1-gerbes with connections, on the one hand, and their holonomies, for simply connected manifolds, or their parallel transports, in the general case, on the other hand. This result is a higher-order analogue of the familiar equivalence between bundles wi…
New algorithm reduces dimensionality in federated learning.
problem Estimating central dimension reduction subspace and variable selection in federated learning.
method Federated sparse sliced inverse regression, convex optimization, linearized alternating direction method of multipliers.
result Upper bound of statistical error rate established under heterogeneous setting.
A new approach combines prior knowledge with learning to adapt quickly to new tasks.
problem Adapting quickly to new tasks using prior knowledge.
method Combines behavior prior, robust off-policy learning, and value function representation.
result Achieves competitive adaptation performance compared to meta reinforcement learning baselines.
In this article, we derive concentration inequalities for the cross-validation estimate of the generalization error for stable predictors in the context of risk assessment. The notion of stability has been first introduced by \cite{DEWA79} and extended by \cite{KEA95}, \cite{BE01} and \cite{KUNIY02} to characterize cla…
New approach connects reinforcement learning robustness and regularization.
problem Dealing with external uncertainty in reinforcement learning.
method Introducing Wasserstein distributionally robust MDPs and a new regularizer.
result Established a dual relation between robust MDPs and regularization.
A classical problem in causal inference is that of matching, where treatment units need to be matched to control units based on covariate information. In this work, we propose a method that computes high quality almost-exact matches for high-dimensional categorical datasets. This method, called FLAME (Fast Large-scale …
In this article, we derive concentration inequalities for the cross-validation estimate of the generalization error for empirical risk minimizers. In the general setting, we prove sanity-check bounds in the spirit of \cite{KR99} \textquotedblleft\textit{bounds showing that the worst-case error of this estimate is not m…
Efficient method improves neural network device placement.
problem Challenging task of placing operations on suitable devices in neural networks.
method End-to-end scalable sequential attention mechanism over graph neural network.
result 16% improvement over human experts and 9.2% over prior art.
Automates OD model selection via meta-learning.
problem Selecting effective OD models for new datasets.
method Meta-learning approach based on past model performances.
result MetaOD selects models significantly outperforming existing methods.
1-D CNNs classify pupil size variations in scotopic conditions.
problem Handling wide inter-subjects variability in pupil analysis.
method 1-D Convolutional Neural Networks applied directly to raw pupil size data.
result 1-D CNNs provide high accuracy in classifying short-range pupil size sequences.
The study sets lower bounds on MMSE for inferring sensitive features from noisy data.
problem Estimating sensitive features from noisy observations of correlated features.
method Adversarial evaluation framework based on MMSE estimation with theoretical lower bounds.
result Derives closed-form bounds for linear models, showing optimality in noise variance.
Enhances polynomial chaos models with uncertainty intervals.
problem Uncertainty quantification in surrogate models.
method Jackknife-based conformal prediction integrated into polynomial chaos expansions.
result Produces accurate predictive intervals for low-accuracy models.
Paper introduces Modular Jets for diagnosing model decompositions in pipelines.
problem Evaluating model decompositions in pipelines for unique identification.
method Estimates empirical jets from module-level representations to diagnose mirage vs identifiable decompositions.
result Proves jet-identifiability theorem for two-module linear regression pipelines.
SAVeD detects dataset versions without metadata, improving accuracy and separation.
problem Difficulty in identifying similar versions of structured datasets.
method Contrastive learning with modified SimCLR pipeline, generating and contrasting augmented table views.
result SAVeD achieves higher accuracy and separation scores on unseen tables.
This paper reviews and advocates against the use of permute-and-predict (PaP) methods for interpreting black box functions. Methods such as the variable importance measures proposed for random forests, partial dependence plots, and individual conditional expectation plots remain popular because they are both model-agno…
Stabilizes online learning by using weighted reservoir sampling.
problem Real-world deployment sensitivity to outliers causes low accuracy in final solutions.
method Weighted reservoir sampling to stabilize ensemble model without additional data passes.
result Risk of ensemble classifier is bounded with respect to the underlying online learning method's regret.
Paper presents a novel gradient-based method for training models and hyperparameters simultaneously.
problem Achieving generalization in machine learning models.
method A novel gradient-based framework that trains parameters and hyperparameters simultaneously.
result Significantly smaller runtime compared to benchmark methods for equivalent prediction scores.
Paper introduces a novel approach to generalize models without validation data.
problem Achieving generalization in machine learning without validation data.
method Introduces a new risk measure and a novel criterion for training models.
result Demonstrates that the new approach outperforms popular methods in generalization and estimation tasks.
A method for non-parametric conditional distribution estimation using CRPS-optimal binning.
problem Non-parametric conditional distribution estimation.
method Partitioning covariate-sorted observations into bins to minimize LOO-CRPS, selecting K by K-fold cross-validation of test CRPS.
result Produces narrower prediction intervals with near-nominal coverage compared to split-conformal competitors.
A new method using negative-shifted gradient descent improves overparameterized linear regression by avoiding structural limitations of negative ridge endpoints.
problem Structural limitations of negative ridge endpoints in overparameterized linear regression.
method Negative-shifted gradient descent, which avoids the pole constraint of negative ridge endpoints.
result The method improves over all admissible endpoints by a polynomial factor in risk under explicit conditions.
Paper presents new training methods for neural networks with tighter risk certificates.
problem Training probabilistic neural networks with tighter risk certificates.
method Derived from PAC-Bayes bounds, two training objectives implemented for the first time in neural networks.
result Competitive test set errors and non-vacuous risk bounds with tighter values than previous results.
Proposes a new framework for evaluating diagnostic models with multiple co-primary endpoints.
problem Overoptimistic assessments of predictive performance in automated medical testing devices.
method Multiple testing framework for diagnostic accuracy studies with co-primary endpoints, using a parametric simultaneous test procedure and Bayesian approach to determine optimal number of models.
result Our approach leads to a better final diagnostic model and increased statistical power.
Develops an algorithm to approximate non-Gaussian posterior distributions in Bayesian inference.
problem Sampling non-Gaussian posterior distributions in Bayesian inverse problems.
method Iterative construction of Gaussian Process (GP) augmented proposal distributions for MCMC sampling.
result Optimal selection of sampling points using maximum information gain from GP surface.
New deep learning model estimates scattering timescale of FRBs efficiently.
problem Estimating scattering timescale of fast radio bursts (FRBs) is a bottleneck.
method Multimodal Transformer Based Generic Mixture Density Network (MT-GMDN) that ingests dynamic spectrum and timeseries profile.
result Achieves 94% R2 on expected value of τ for measurable scattering. Incremental machine learning models predict COVID-19 cases more efficiently than traditional methods.
problem Predicting the spread of COVID-19 cases in real-time across multiple countries.
method Comparison of online incremental machine learning algorithms against traditional LSTM models.
result Incremental machine learning models are more efficient and computationally cheaper than traditional methods.
Though deep reinforcement learning has led to breakthroughs in many difficult domains, these successes have required an ever-increasing number of samples. As state-of-the-art reinforcement learning (RL) systems require an exponentially increasing number of samples, their development is restricted to a continually shrin…