This paper addresses overfitting in dimension reduction methods by calibrating hyperparameters considering noise.
problem Overfitting in dimension reduction methods, especially t-SNE and UMAP, when data contains noise.
method Present a framework to calibrate hyperparameters in the presence of noise for t-SNE and UMAP.
result Recommended hyperparameter values for t-SNE and UMAP are too small and overfit the noise.
This study improves hyperparameter optimization for categorical and non-normal data.
problem Bayesian hyperparameter optimization struggles with categorical hyperparameters and non-normal data.
method Integrates conformalized quantile regression to address estimation weaknesses and provides robust calibration guarantees.
result Quantile surrogate architectures and acquisition functions yield superior performance compared to existing methods.
Paper proposes HCDC to improve hyperparameter search efficiency.
problem Poor generalizability of dataset condensation across different hyperparameters.
method HCDC algorithm that matches hyperparameter gradients for synthetic validation dataset.
result HCDC effectively maintains validation-performance rankings of models.
Unified Bayesian-AI framework improves epidemiological risk prediction and uncertainty quantification.
problem Lack of calibrated uncertainty in machine learning models for epidemiology.
method Combines Bayesian prediction with Bayesian hyperparameter optimization using logistic regression and Gaussian-process Bayesian optimization.
result Unified Bayesian-AI framework provides reliable coverage and improved calibration, enhancing epidemiological decision making.
This paper calibrates Gaussian process predictive distributions for Bayesian optimization to improve sampling decisions.
problem Lower-tail miscalibration in GP predictive distributions affects BO sampling decisions.
method Introduces goal-oriented calibration for GP predictive distributions below a threshold t. result Post-hoc method tcGP improves lower-tail calibration and BO performance.
A new method for effective VAE training using calibrated decoders.
problem Training VAEs requires hyperparameter tuning, leading to inefficiency.
method Calibrated decoders that learn uncertainty and automatically determine information retention.
result Calibrated decoders can simplify VAE training without heuristic modifications.
TCP provides well-calibrated prediction intervals for nonstationary time series.
problem Nonstationary time series forecasting with well-calibrated prediction intervals.
method Temporal Conformal Prediction (TCP) couples a modern quantile forecaster with a rolling split-conformal calibration layer.
result TCP achieves near-nominal coverage, providing slightly wider intervals than Historical Simulation.
Develops a power-calibrated framework for LLM watermarking, optimizing tradeoffs between detectability and distortion.
problem The trade-off between detectability and semantic distortion in logit-based watermarking.
method Power-calibrated statistical framework for watermark hyperparameters, establishing explicit relationships.
result Derives practical parameter selection procedures achieving optimal tradeoffs under constraints.
A cubing strategy identifies stable hyperparameter regions for uncertainty quantification in spatial deep learning.
problem Uncertainty quantification in spatial deep learning models.
method Cubing-based diagnostic framework to recursively partition hyperparameter space and evaluate regions using scoring rules.
result Our approach produces competitive or superior predictive intervals compared to a statistical baseline model.
A new method calibrates Gaussian processes for more accurate uncertainty estimates.
problem Uncertainty estimates from Gaussian processes are often miscalibrated in practice.
method A novel calibration approach using different hyperparameters to generate more accurate predictive quantiles.
result The method yields tighter predictive quantiles and is more flexible than existing approaches.
The paper proposes a new method to calibrate option pricing models that accurately match both volatility surfaces and variance term structures.
problem Calibrated models often produce inaccurate variance term structures relative to market observations.
method The paper introduces a joint calibration framework that augments the conventional objective function with a penalty term for variance term structure deviations, using a hyperparameter to balance volatility surface and variance term structure weights.
result The proposed method accurately fits observed option prices while delivering realistic term structures of variance.
Optimizes calibration error estimators for better classifier trustworthiness.
problem Lack of guidance on selecting and tuning calibration error estimators.
method Reformulates calibration estimation as a regression problem with i.i.d. input pairs.
result Demonstrates the effectiveness of optimized calibration estimators on image classification tasks.
Probability estimates generated by boosting ensembles are poorly calibrated because of the margin maximization nature of the algorithm. The outputs of the ensemble need to be properly calibrated before they can be used as probability estimates. In this work, we demonstrate that online boosting is also prone to producin…
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.
Paper develops an efficient approach to reduce HPO time.
problem Challenges in determining optimal hyperparameters due to large number and training time.
method Nested Latin hypercube design for initialization, truncated additive Gaussian process model for calibration, sequential model-based algorithm for optimization.
result Demonstrates competitive performance on various machine learning models.
New hyperparameter ensembles boost neural network performance and uncertainty.
problem Improving neural network robustness and uncertainty quantification.
method Designing ensembles over both weights and hyperparameters, stratified across random initializations.
result Hyper-deep and hyper-batch ensembles outperform deep and batch ensembles on various architectures.
The paper improves Bayesian optimization by calibrating uncertainty estimates.
problem Improper uncertainty estimates in Bayesian optimization when data is non-stationary.
method Proposes online learning algorithms to maintain calibration on non-i.i.d. data and integrates them into Bayesian optimization.
result Calibrated Bayesian optimization converges to better optima in fewer steps.
TransCal calibrates DA models with lower bias and variance.
problem Calibrating DA models to estimate accurate predictive uncertainty.
method Transferable Calibration (TransCal) in a unified hyperparameter-free optimization framework.
result TransCal achieves more accurate calibration with lower bias and variance.
SVI and GP surrogates improve calibration of ABMs in epidemiology.
problem Calibrating stochastic ABMs in epidemiology is computationally expensive.
method Stein Variational Inference (SVI) with Gaussian process (GP) surrogates.
result SVI maintains comparable predictive accuracy and calibration effectiveness to MCMC.
Focal loss improves deep neural networks' accuracy and calibration.
problem Miscalibration in deep neural networks.
method Using focal loss and temperature scaling to improve model calibration.
result Focal loss leads to state-of-the-art calibrated models without sacrificing accuracy.
OptFormer learns universal HPO from diverse datasets.
problem Learning HPO from experiments with different hyperparameters.
method Text-based Transformer framework for joint policy and function prediction.
result OptFormer can imitate multiple HPO algorithms and improve predictions.
End-to-end method improves neural network calibration during training.
problem Improving neural network calibration for regression problems.
method Quantile Recalibration Training integrates post-hoc calibration into model training.
result Improved predictive accuracy and calibration in a large-scale experiment.
Study improves neural network calibration for drug discovery.
problem Improper calibration of neural network predictions in drug discovery.
method Compared different metrics for model hyperparameter tuning and proposed Bayesian Linear Probing (BLP) method.
result Bayesian Linear Probing (BLP) improves model calibration and accuracy.
The performance of deep neural networks is highly sensitive to the choice of the hyperparameters that define the structure of the network and the learning process. When facing a new application, tuning a deep neural network is a tedious and time consuming process that is often described as a "dark art". This explains t…
Online Platt Scaling adapts to varying data distributions.
problem Adapting Platt scaling to non-i.i.d. settings with distribution drift.
method Combines Platt scaling with online logistic regression and calibeating.
result OPS+calibeating method is guaranteed to be calibrated for adversarial outcomes.
New method estimates marginal likelihood for deep learning models using training data alone.
problem Estimation difficulties in marginal likelihood for model selection in deep learning.
method Scalable marginal likelihood estimation based on Laplace's method and Gauss-Newton approximations.
result Estimate outperforms cross-validation and manual tuning on various datasets.
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.
Study optimizes tree-based models for better alignment of predicted scores and actual probabilities.
problem Traditional calibration metrics fail to align predicted scores with actual probabilities when score distributions deviate from the underlying data.
method Optimizes tree-based models (Random Forest, XGBoost) using Kullback-Leibler (KL) divergence to minimize the difference between predicted and true probability distributions.
result Optimized tree-based models yield superior alignment between predicted scores and actual probabilities without significant performance loss.
New methods for distributed CP improve reliability in healthcare.
problem Post-hoc calibration of models in safety-critical domains.
method Quantile-based and histogram-based distributed conformal prediction via message passing.
result Improved reliability and coverage guarantees in decentralized settings.
This paper enhances the Random Survival Forest model for better predictive maintenance.
problem Improving time-to-failure estimation in predictive maintenance with Random Survival Forest.
method A three-level framework for quantifying hyperparameter tunability, including model-level and hyperparameter-level metrics.
result Hyperparameter tuning significantly improves Random Survival Forest model performance in predictive maintenance.
We aim to design adaptive online learning algorithms that take advantage of any special structure that might be present in the learning task at hand, with as little manual tuning by the user as possible. A fundamental obstacle that comes up in the design of such adaptive algorithms is to calibrate a so-called step-size…
It is important to identify the change point of a system's health status, which usually signifies an incipient fault under development. The One-Class Support Vector Machine (OC-SVM) is a popular machine learning model for anomaly detection and hence could be used for identifying change points; however, it is sometimes …
A new method for kernel tests without data splitting increases power.
problem Lack of power in kernel-based tests due to data splitting.
method Selective inference framework to learn hyperparameters and test on full sample.
result Empirically larger test power without data splitting, regardless of split proportion.
Deterministic neural networks (NNs) are increasingly being deployed in safety critical domains, where calibrated, robust, and efficient measures of uncertainty are crucial. In this paper, we propose a novel method for training non-Bayesian NNs to estimate a continuous target as well as its associated evidence in order …
A general approach for anomaly detection or novelty detection consists in estimating high density regions or Minimum Volume (MV) sets. The One-Class Support Vector Machine (OCSVM) is a state-of-the-art algorithm for estimating such regions from high dimensional data. Yet it suffers from practical limitations. When appl…
Deep heteroskedastic models overfit, showing a phase transition with regularization strength.
problem Overfitting in deep heteroskedastic regression models.
method Theoretical framework based on statistical field theory, empirical verification, and hyperparameter simplification.
result A phase transition in model behavior with varying regularization strength.
Hybrid model outperforms benchmarks in financial forecasting.
problem Robust asset price forecasting in finance.
method Combining LSTM with Neural Levy Processes using Grey Wolf Optimizer and ANN calibration.
result Hybrid model outperforms base LSTM and other models.
We introduce a tunable loss function called α-loss, parameterized by α∈(0,∞], which interpolates between the exponential loss (α=1/2), the log-loss (α=1), and the 0-1 loss (α=∞), for the machine learning setting of classification. Theoretically, we illustrate a fundamental connection between $…
New method detects changes in high-dimensional data from small samples.
problem Detecting changes in high-dimensional data with limited samples.
method Angular kernel scan framework for detecting marginal distributional shifts.
result Exact population mean factorization and asymptotically distribution-free test.
Adapts safe policies for exploration in high-risk settings.
problem Balancing safety and exploration in high-risk environments.
method Uses conformal calibration on a safe reference policy to determine aggressive action limits.
result Safe exploration improves performance without requiring model class identification or hyperparameter tuning.
DPConvCNP learns to predict private data accurately and privately.
problem Balancing privacy and accuracy in machine learning models.
method Meta-learning combined with improved DP mechanism.
result DPConvCNP outperforms DP GP baseline, especially on non-Gaussian data.
The paper shows cross-validation fails in learning Gaussian graphical model structures.
problem Cross-validation's failure in learning Gaussian graphical model structures.
method Finite-sample bounds on misidentification probability of Lasso estimator.
result Cross-validation is inconsistent for learning Gaussian graphical model structures.
SurvivalPFN simplifies survival analysis through amortized Bayesian inference.
problem Selecting appropriate survival analysis methods requires expertise and can be time-consuming.
method SurvivalPFN uses a prior-data fitted network for in-context Bayesian inference.
result SurvivalPFN achieves strong predictive performance across diverse datasets.
Proposes a new trust framework for AI models to maximize utility.
problem Concerns over bias and discrimination in predictive models.
method Introduces a novel trust framework inspired by philosophy, focusing on maximizing Bayes utility.
result Properly-ranked models are inherently U-trustworthy. Hyperparameter optimization can be formulated as a bilevel optimization problem, where the optimal parameters on the training set depend on the hyperparameters. We aim to adapt regularization hyperparameters for neural networks by fitting compact approximations to the best-response function, which maps hyperparameters …
This work analyzes aggregation strategies for Bayesian deep learning models in federated learning.
problem Improper aggregation of Bayesian deep learning models in federated learning leads to sub-optimal performance.
method Six aggregation strategies for Bayesian deep learning models are analyzed using CIFAR-10 dataset and a fully variational ResNet-20 architecture.
result Aggregation strategy is a key hyperparameter affecting accuracy, calibration, uncertainty quantification, training stability, and client compute requirements.
Generative networks minimize predictive scoring rules for probabilistic forecasting.
problem Evaluating and improving probabilistic forecasts using generative models.
method Training generative networks to minimize predictive-sequential scoring rules on temporal sequences.
result Our method outperforms adversarial approaches in probabilistic calibration.
Deep Gaussian Processes (DGPs) are hierarchical generalizations of Gaussian Processes that combine well calibrated uncertainty estimates with the high flexibility of multilayer models. One of the biggest challenges with these models is that exact inference is intractable. The current state-of-the-art inference method, …