mlsanityData to 7 Oct 2026

Confidence-based Estimators for Predictive Performance in Model Monitoring

arXiv 2407.08649 ↗First posted 11 Jul 2024cs.LG cs.AIcited by 8

Not on the map. Why?

Abstract (arXiv)

After a machine learning model has been deployed into production, its predictive performance needs to be monitored. Ideally, such monitoring can be carried out by comparing the model's predictions against ground truth labels. For this to be possible, the ground truth labels must be available relatively soon after inference. However, there are many use cases where ground truth labels are available only after a significant delay, or in the worst case, not at all. In such cases, directly monitoring the model's predictive performance is impossible. Recently, novel methods for estimating the predictive performance of a model when ground truth is unavailable have been developed. Many of these methods leverage model confidence or other uncertainty estimates and are experimentally compared against a naive baseline method, namely Average Confidence (AC), which estimates model accuracy as the average of confidence scores for a given set of predictions. However, until now the theoretical properties of the AC method have not been properly explored. In this paper, we try to fill this gap by reviewing the AC method and show that under certain general assumptions, it is an unbiased and consistent estimator of model accuracy with many desirable properties. We also compare this baseline estimator against some more complex estimators empirically and show that in many cases the AC method is able to beat the others, although the comparative quality of the different estimators is heavily case-dependent.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 34 of its 52 references that are in the map. First posted on arXiv 11 Jul 2024.

Builds on

34 of 52 references.

References · 34

  1. 2017On Calibration of Modern Neural Networkscited by 4,860
  2. 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,362
  3. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
  4. 2019Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shiftcited by 1,274
  5. 2019When Does Label Smoothing Help?cited by 985
  6. 2019Measuring Calibration in Deep Learningcited by 443
  7. 2016Kernel Mean Embedding of Distributions: A Review and Beyondcited by 360
  8. 2019Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibrationcited by 359
  9. 2020Calibrating Deep Neural Networks using Focal Losscited by 353
  10. 2019Verified Uncertainty Calibrationcited by 307
  11. 2018Learning Confidence for Out-of-Distribution Detection in Neural Networkscited by 245
  12. 2018Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shiftcited by 240
22 more
  1. 2020Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learningcited by 190
  2. 2019Addressing Failure Prediction by Learning Model Confidencecited by 183
  3. 2019Evaluating model calibration in classificationcited by 165
  4. 2022Leveraging Unlabeled Data to Predict Out-of-Distribution Performancecited by 141
  5. 2021Assessing Generalization of SGD via Disagreementcited by 116
  6. 2021Predicting with Confidence on Unseen Distributionscited by 116
  7. 2022Agreement-on-the-Line: Predicting the Performance of Neural Networks under Distribution Shiftcited by 94
  8. 2022From Concept Drift to Model Degradation: An Overview on Performance-Aware Drift Detectorscited by 82
  9. 2023Benchmarking Distribution Shift in Tabular Data with TableShiftcited by 64
  10. 2021Mandoline: Model Evaluation under Distribution Shiftcited by 55
  11. 2021Detecting Errors and Estimating Accuracy on Unlabeled Data with Self-training Ensemblescited by 54
  12. 2018Learning for Single-Shot Confidence Calibration in Deep Neural Networks through Stochastic Inferencescited by 25
  13. 2021Estimating Expected Calibration Errorscited by 24
  14. 2020Monitoring and explainability of models in productioncited by 23
  15. 2023Confidence and Dispersity Speak: Characterising Prediction Matrix for Unsupervised Accuracy Estimationcited by 22
  16. 2018Confidence Scoring Using Whitebox Meta-models with Linear Classifier Probescited by 21
  17. 2021Detecting Concept Drift With Neural Network Model Uncertaintycited by 21
  18. 2023Characterizing Out-of-Distribution Error via Optimal Transportcited by 20
  19. 2021MLDemon: Deployment Monitoring for Machine Learning Systemscited by 14
  20. 2021Rethinking Importance Weighting for Transfer Learningcited by 11
  21. 2024Estimating Model Performance Under Covariate Shift Without Labelscited by 9
  22. 2020Calibrating Deep Neural Network Classifiers on Out-of-Distribution Datasetscited by 7

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 4; 2026: 3.

8 papers.

  1. 2025Suitability Filter: A Statistical Framework for Classifier Evaluation in Real-World Deployment Settingscited by 4
  2. 2025Performance Estimation in Binary Classification Using Calibrated Confidencecited by 2
  3. 2024Predictive maintenance solution for industrial systems -- an unsupervised approach based on log periodic power lawcited by 0
  4. 2025Unsupervised Source-Free Ranking of Biomedical Segmentation Models Under Distribution Shiftcited by 0
  5. 2025Label-free estimation of clinically relevant performance metrics under distribution shiftscited by 0
  6. 2026CaliPPer: quantifying, predicting and improving AI model performance for binding predictioncited by 0
  7. 2026Regime-Conditional Verification: Correctness Estimation for Adapting and Monitoring Safety Classifierscited by 0
  8. 2026FMMO: Detecting the Divergence Between Local Attribution and Global Driftcited by 0