Post-hoc transforms can reverse model performance trends, especially in noisy settings.
problem Post-hoc transforms can reverse model performance trends, especially in noisy settings.
method Empirical study and analysis of post-hoc transforms like temperature scaling, ensembling, and SWA.
result Post-hoc reversal can prevent double descent and mitigate mismatches between test loss and test error.
While statistics and machine learning offers numerous methods for ensuring generalization, these methods often fail in the presence of adaptivity---the common practice in which the choice of analysis depends on previous interactions with the same dataset. A recent line of work has introduced powerful, general purpose a…
PH-CS selects test inputs with reliability guarantees, adapting FDR to data.
problem Fixed FDR limits adaptability to downstream needs.
method Post-hoc conformal selection with e-variables, e-BH procedure.
result PH-CS provides reliable FDP estimates and competitive FDR control.
EAGLE improves reproducibility and stability of model explanations.
problem Creating reliable explanations for opaque machine learning models.
method Formulates perturbation selection as an information-theoretic active learning problem.
result EAGLE learns a linear surrogate model with feature importance scores and uncertainty estimates.
The paper provides high-probability bounds on false discovery proportions in conformal inference.
problem Existing methods fail to provide high-probability bounds on the realized false discovery proportion.
method Constructing a high-probability envelope for the empirical distribution function of null conformal p-values by sampling from their joint distribution.
result Establishes finite-sample, distribution-free upper bounds on the FDP that hold simultaneously over all possible rejection thresholds.
iKF method uncovers complex variable interactions for scientific discovery.
problem Limited interpretability of existing models in decision-making applications.
method Iterative Kings' Forests (iKF) method to uncover multi-order interactions.
result iKF provides strong interpretive power for explainable modeling.
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.
Paper proposes hybrid approach for transparent credit scoring models.
problem Lack of transparency in machine learning models limits their use in regulated environments.
method Post-hoc interpretation of black-box models guides feature selection, followed by training glass-box models.
result Reduces feature usage from 106 to 10 while maintaining comparable performance.
Post-hoc calibration improves uncertainty under domain shift.
problem Improving uncertainty calibration under domain shift.
method Apply perturbations to validation set before post-hoc calibration.
result Perturbation step results in better calibration under domain shift.
Framework evaluates post-hoc interpretability methods in time-series classification.
problem Lack of suitable post-hoc interpretability methods for time-series classification.
method Proposes a framework with quantitative metrics to assess interpretability methods.
result Addresses several drawbacks of existing methods, including dependence on human judgement and data distribution shift.
Proposes a method for explaining tabular data using copulas.
problem Lack of ground truth for explainability in complex datasets.
method Uses copulas to specify statistical properties and build intuition.
result Demonstrates improved explainability on logistic regression and correlation use cases.
A framework for quantifying uncertainty in feature importance values.
problem Stable interpretation of feature importance values in machine learning models.
method A novel method based on pairwise comparisons of feature importance values to produce confidence intervals for feature ranks.
result The method produces simultaneous confidence intervals for feature ranks, enabling selection of top-k important features.
Mathematical study shows post-hoc explanations are better than attention weights alone.
problem Understanding the internal behavior of attention-based models.
method Mathematical analysis of a simple attention-based architecture.
result Post-hoc explanations provide more useful insights than attention weights alone.
PACE explains ViTs by modeling patch-level concept distributions, surpassing existing methods.
problem Lack of trustworthy post-hoc explanations for Vision Transformers (ViTs)
method Variational Bayesian explanation framework (PACE)
result PACE surpasses state-of-the-art methods in meeting desiderata for ViT explanations.
CalArena benchmarks post-hoc calibration methods across various tasks.
problem Inconsistent evaluations of post-hoc calibration methods.
method Large-scale benchmark with 2000 experiments, covering diverse models and settings.
result Smooth calibration functions outperform binning-based approaches.
Meta-Cal improves post-hoc calibration of neural networks.
problem Improving the accuracy of uncalibrated neural network predictions.
method Meta-Cal uses a base calibrator and a ranking model with constraints to provide high-probability bounds.
result Meta-Cal significantly outperforms existing methods in post-hoc multi-class classification calibration.
Study post-hoc Learning to Defer using density-ratio losses.
problem Optimizing decision-making between models and experts.
method Density-ratio losses for post-hoc L2D scorers, derived from class-probability estimation.
result The approach recovers known results and introduces new connections to expert comparison and anomaly detection.
Post-hoc calibration of neural networks using g-Layers proves theoretical justification.
problem Ensuring the confidence of neural network decisions in real-world applications.
method Proves theoretical justification for post-hoc calibration methods by adding g-Layers and minimizing NLL.
result Proves that adding g-Layers and minimizing NLL can lead to a calibrated network.
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.
A post-hoc framework improves model performance by calibrating different feature spaces.
problem Improving AUC performance on binary classification tasks for overconfident models.
method Identifies heterogeneous partitions of the feature space and applies post-hoc calibration techniques to each partition.
result Theoretical optimality of the framework for any model, demonstrated on deep neural networks.
Counterfactual post-hoc interpretability approaches have been proven to be useful tools to generate explanations for the predictions of a trained blackbox classifier. However, the assumptions they make about the data and the classifier make them unreliable in many contexts. In this paper, we discuss three desirable pro…
Optimal selective classification using likelihood ratios improves model reliability.
problem Enhancing predictive model reliability by allowing uncertain predictions.
method Neyman--Pearson lemma applied to likelihood ratios for optimal selection.
result Neyman--Pearson-informed methods outperform existing baselines under covariate shifts.
A RL framework selects features to balance bias and accuracy dynamically.
problem Bias in automated feature selection when predictors are correlated.
method Multi-component reward function with policy gradient for dynamic regularization and bias mitigation.
result Model balances fairness and accuracy during training.
This review explores methods to explain deep neural networks and their applications.
problem Understanding the decision-making process of deep neural networks.
method Overview of interpretability methods, theoretical foundations, and comparative evaluations.
result Demonstrates the effectiveness of explainable AI in various applications.
Greedy PIG adapts integrated gradients for better feature attribution.
problem Interpreting deep learning model predictions.
method Unified discrete optimization framework for feature attribution and selection.
result Greedy PIG improves feature attribution on various tasks.
RENT selects stable features for robust model interpretation.
problem Stability in feature selection for robust model interpretation.
method RENT uses an ensemble of elastic net models trained on distinct subsets of data.
result RENT selects features with high stability improving model robustness.
Selection of input features such as relevant pieces of text has become a common technique of highlighting how complex neural predictors operate. The selection can be optimized post-hoc for trained models or incorporated directly into the method itself (self-explaining). However, an overall selection does not properly c…
New method calibrates deep networks by preserving top-k predictions.
problem Calibrated confidence scores for multi-class deep networks to avoid rare mistakes.
method Intra order-preserving functions combined with neural network architecture.
result Outperforms state-of-the-art methods in evaluation metrics.
Post-hoc explanations improve CNNs by replacing final linear layer with k-means classifier.
problem CNNs lack accurate data representation in their built-in prototypes.
method Introduces k-means-based post-hoc explanations for CNNs, leveraging spatial consistency of convolutional receptive fields.
result Using shallower, less compressed feature activations improves semantic fidelity at the cost of slight predictive performance.
A novel post-hoc calibration method reduces neural network calibration errors.
problem Neural networks produce poorly calibrated probabilities, leading to underconfidence and overconfidence.
method Probability bounding (PB) via box-constrained softmax (BCSoftmax) function.
result Consistently reduces calibration errors on four real-world datasets.
TaylorPODA uses Taylor expansions to improve feature attributions for opaque models.
problem Lack of systematic framework for quantifying feature contributions in opaque models.
method Taylor expansion framework with postulates (precision, federation, zero-discrepancy, adaptation).
result TaylorPODA achieves competitive results and provides principled explanations.
Post-hoc model-agnostic interpretation methods such as partial dependence plots can be employed to interpret complex machine learning models. While these interpretation methods can be applied regardless of model complexity, they can produce misleading and verbose results if the model is too complex, especially w.r.t. f…
Post-detection analysis identifies responsible coordinates for multivariate change-points.
problem Identifying which coordinates in multivariate time series change after a detected change-point.
method Two-sample testing procedures with nonparametric tests for Type I error control.
result Strong performance of proposed post hoc statistical procedures.
CUBE explains models by balanced experiments and contrasts.
problem Post-hoc explanation of trained predictive models.
method Design-based framework using balanced low-high probes.
result Reveals dominant learned effect structure and clarifies query efficiency.
New analysis reveals gaps in selective classifiers, guiding improvements.
problem Improving selective classifiers to match perfect-ordering oracle performance.
method Formalized selective classification gap, decomposed into five sources of looseness.
result Monotone post-hoc calibration has limited impact on closing the gap.
Conformal Bayes under label shift: post-hoc calibration vs. in-training adaptation
problem Bayesian prediction sets under label shift
method Post-hoc calibration vs. In-training adaptation
result Both strategies achieve valid coverage equally in an unbiased training regime
A new feature selection method for cost-sensitive classification in Random Forests.
problem Feature selection in Random Forests is challenging due to the complexity of ensemble trees.
method Shallow Tree Selection method for feature selection from small tree structures, adapted for cost-sensitive learning.
result BCR criterion improves performance of feature selection methods.
New algorithm reduces adaptation lag in online model selection.
problem Adaptation lag in online model selection for non-stationary environments.
method Optimistic online mirror descent with safeguarded large learning rates.
result Reduces adaptation lag from hundreds of rounds to a few rounds.
New asymptotic e-values improve inference by eliminating data-dependent scaling inefficiency.
problem Data-dependent scaling inefficiency in existing asymptotic e-values.
method Drawing on Bentkus's near-optimal concentration inequalities, introduce Bentkus-type asymptotic e-values.
result Bentkus-type asymptotic e-values consistently deliver sharper inference than existing alternatives.
DFI maps covariates to latent representations for feature importance.
problem Feature importance when predictors are statistically dependent.
method Disentangled Feature Importance (DFI) using entropic optimal transport.
result DFI yields stable, interpretable, uncertainty-quantified attributions of shared predictive signal.
Post-hoc uncertainty quantification improves on pre-trained neural networks without underfitting.
problem Uncertainty quantification in neural networks is underfitting or computationally demanding.
method Gaussian Process Activation function (GAPA) for neuron-level uncertainty, with two methods: GAPA-Free and GAPA-Variational.
result GAPA-Variational outperforms Laplace approximation on most datasets in uncertainty quantification metrics.
Improved probabilistic solar irradiance forecasting models for grid integration.
problem Enhancing accuracy of solar irradiance forecasts for grid integration.
method Developed and calibrated probabilistic models using post-hoc calibration techniques.
result NGBoost model with CRUDE calibration achieves comparable performance to numerical weather prediction models.
We describe a post hoc test for the Sharpe ratio, analogous to Tukey's test for pairwise equality of means. The test can be applied after rejection of the hypothesis that all population Signal-Noise ratios are equal. The test is applicable under a simple correlation structure among asset returns. Simulations indicate t…
New method calibrates neural network predictions for better reliability.
problem Improper probability estimates from deep networks leading to unreliable predictions.
method Proposes a constrained optimization approach for a monotonic calibration map.
result Achieves state-of-the-art performance across various datasets and models.
Two approaches improve conformal Bayes for label shift, one post-hoc and one in-training.
problem Improving prediction sets for target domain under label shift.
method Two complementary approaches: post-hoc calibration and in-training adaptation.
result In-training adaptation achieves up to 43% width reduction at unchanged coverage.
Combines Laplace approximations of deep networks for better uncertainty quantification.
problem Overconfident predictions on outliers in deep learning models.
method Gaussian mixture model posterior using weighted sum of Laplace approximations of pre-trained deep networks.
result Mitigates overconfidence 'far away' from training data.
Enhances deep neural networks with fixed-mean Gaussian processes for uncertainty estimation.
problem Post-hoc uncertainty estimation of pre-trained deep neural networks.
method Fixed-mean Gaussian processes with variational inference for efficient stochastic optimization.
result FMGP improves uncertainty estimation and computational efficiency compared to state-of-the-art methods.
The paper explores how model complexity affects OOD detection performance.
problem Ensuring reliability and safety of machine learning systems through OOD detection.
method Investigates the relationship between model capacity and OOD detection performance using empirical and theoretical analysis.
result The Double Descent phenomenon is observed in post-hoc OOD detection, indicating that overparameterization can enhance OOD detection.