ManifoldMind uses adaptive-curvature probabilistic spheres for trustworthy recommendations in semantic hierarchies.
problem Sparse and abstract recommendation domains where users explore diverse conceptual paths.
method Adaptive-curvature probabilistic spheres, soft multi-hop inference, and curvature-aware semantic kernel.
result Superior NDCG, calibration, and diversity compared to baselines on public benchmarks.
User strategization undermines algorithmic trustworthiness.
problem User strategic behavior corrupts algorithmic data and trust.
method Modeling user-platform interactions as a game, analyzing strategic behavior's short-term benefits and long-term harms.
result User strategization can initially benefit platforms but ultimately harms their ability to make accurate decisions.
SAFER improves personalized treatment recommendations for dynamic clinical contexts.
problem Personalized treatment optimization in evolving clinical contexts with safety concerns.
method Integrates structured EHR and clinical notes, uses conformal prediction for safe recommendations.
result SAFER outperforms state-of-the-art baselines in recommendation metrics and mortality rates.
Clarifies challenges in machine learning uncertainty quantification.
problem Inconsistent terminology and diverse technical requirements for trustworthy uncertainties.
method Examines estimation targets, uncertainty constructs, and problematic mappings.
result Advocates for alignment between intent and implementation in UQ.
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. A framework assesses the trustworthiness of probabilistic classifiers using local calibration error.
problem Assessing the trustworthiness of probabilistic classifiers beyond traditional metrics.
method I-trustworthy framework linking local calibration to trustworthiness; Kernel Local Calibration Error (KLCE) method for hypothesis testing.
result The effectiveness of the proposed test statistic demonstrated through simulated and real-world datasets.
Media seems to have become more partisan, often providing a biased coverage of news catering to the interest of specific groups. It is therefore essential to identify credible information content that provides an objective narrative of an event. News communities such as digg, reddit, or newstrust offer recommendations,…
Neural approach enhances AI trustworthiness, generalization, and robustness.
problem Challenges in explaining, generalizing, and adapting AI models to uncertain environments.
method Customized trustworthy networks, flexible learning regularizers, open-world recognition losses.
result Significant performance improvements across various open-world multimedia recognition scenarios.
PolicySynth improves synthetic data alignment with real data for better campaign decisions.
problem Synthetic data used in decision support systems often leads to incorrect decisions.
method PolicySynth framework that conditions synthetic data on churn scorer to align with real data decisions.
result PolicySynth achieves high strategy simulation fidelity (0.923-0.960) on churn and acquisition datasets.
Bayesian meta learning improves uncertainty quantification in regression.
problem Trusting uncertainty quantification in Bayesian regression.
method Trust-Bayes framework for Bayesian meta learning, optimizing for trustworthy uncertainty quantification.
result Lower bounds and sample complexity for trustworthy uncertainty quantification are characterized.
Online knowledge repositories typically rely on their users or dedicated editors to evaluate the reliability of their content. These evaluations can be viewed as noisy measurements of both information reliability and information source trustworthiness. Can we leverage these noisy evaluations, often biased, to distill a…
A framework for analyzing regularizers to ensure trustworthy theory-driven model estimation.
problem Uncertain choice of regularizers can compromise the interpretability of deep grey-box models.
method Adapting neural net architecture and training objective to analyze regularizer behavior empirically.
result Empirical analysis of regularizers helps in making a justified choice for trustworthy theory-driven model estimation.
AI systems need reliable testing to ensure safety and trustworthiness.
problem Current AI Act lacks functional trustworthiness for AI systems.
method Define technical application distribution, set risk-based performance, and conduct statistically valid testing.
result Reliable functional trustworthiness is essential for AI systems.
New research finds tree-based GAMs are most trustworthy and fair.
problem Variability in GAM algorithms leads to inconsistent models.
method Quantitative and qualitative analysis of various GAM algorithms.
result Tree-based GAMs are the most trustworthy and fair.
Framework audits synthetic datasets for trustworthiness across various use cases.
problem Assessing the trustworthiness of synthetic datasets and models.
method Holistic auditing framework focusing on bias, fidelity, utility, robustness, and privacy.
result Introduces a trustworthiness index and model selection process for controllable trade-offs.
In the context of fake news, bias, and propaganda, we study two important but relatively under-explored problems: (i) trustworthiness estimation (on a 3-point scale) and (ii) political ideology detection (left/right bias on a 7-point scale) of entire news outlets, as opposed to evaluating individual articles. In partic…
Efficient approach improves prediction calibration for domain shifts.
problem Improving uncertainty-aware predictions for domain shifts.
method Combining entropy-encouraging and adversarial calibration losses.
result Substantially outperforms existing approaches in domain drift calibration.
Paper forecasts tax revenues in Bulgaria during pandemic.
problem Forecasting tax revenues during pandemic.
method Model based on IMF recommendations, using 1995-2019 data.
result Pandemic negatively impacts tax revenues, but econometrics can still produce forecasts.
Proposes a game-theoretic framework for ML trust regulation.
problem Lack of coordination between ML model builders and regulators.
method Formulates trustworthy ML as a multi-objective multi-agent optimization problem and introduces regulation games and ParetoPlay.
result Enables efficient enforcement of ML model specifications without discouraging participation.
We propose Top-N-Rank, a novel family of list-wise Learning-to-Rank models for reliably recommending the N top-ranked items. The proposed models optimize a variant of the widely used discounted cumulative gain (DCG) objective function which differs from DCG in two important aspects: (i) It limits the evaluation of DCG …
Survey of technologies for trustworthy machine learning systems.
problem Building machine learning systems that are fair, explainable, auditable, and secure.
method Survey of technologies across data and model stages of machine learning.
result Four categories of system properties (fairness, explainability, auditability, safety & security) are essential for trustworthy systems.
Study shows trust and trustworthiness emerge through reinforcement learning.
problem Trust and trustworthiness are universal but not predicted by traditional economic models.
method Used Q-learning algorithm to simulate trust and trustworthiness dynamics in a trust game.
result High levels of trust and trustworthiness emerge when individuals consider both past and future experiences.
Paper monitors DNN accuracy to enhance trustworthiness.
problem Varying DNN accuracy in practice and lack of ground truth labels.
method Post-hoc accuracy monitor model using Monte-Carlo dropout ensemble.
result Accuracy monitor provides close-to-true accuracy estimation.
While social networks can provide an ideal platform for up-to-date information from individuals across the world, it has also proved to be a place where rumours fester and accidental or deliberate misinformation often emerges. In this article, we aim to support the task of making sense from social media data, and speci…
Online health communities are a valuable source of information for patients and physicians. However, such user-generated resources are often plagued by inaccuracies and misinformation. In this work we propose a method for automatically establishing the credibility of user-generated medical statements and the trustworth…
Conformal Alignment ensures trustworthy outputs from foundation models.
problem Ensuring outputs from foundation models align with human values in high-stakes tasks.
method A framework that trains an alignment predictor using reference data to select trustworthy outputs.
result Conformal Alignment accurately identifies trustworthy outputs via lightweight training over moderate reference data.
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.
Paper explores differential privacy in high-dimensional federated learning, tackling server trustworthiness and estimation.
problem Maintaining privacy in distributed environments with high-dimensional data.
method Investigates scenarios with untrusted and trusted central servers, introduces novel federated estimation algorithms for linear regression models.
result Tight minimax rates depend on high-dimensionality even with sparsity assumptions, and novel algorithms handle slight variations among distributed models.
New trust matrix quantifies breakdowns in deep neural networks.
problem Understanding trust breakdowns in deep learning models.
method Introduces trust matrix and conditional trust densities to analyze deep neural networks.
result Trust matrices reveal areas needing improvement for deep neural networks.
Despite the great success achieved in machine learning (ML), adversarial examples have caused concerns with regards to its trustworthiness: A small perturbation of an input results in an arbitrary failure of an otherwise seemingly well-trained ML model. While studies are being conducted to discover the intrinsic proper…
Consensus dimension reduction combines multiple visualizations to identify shared patterns.
problem Conflicting visualizations from different dimension reduction methods.
method Multi-view learning to identify stable patterns across multiple views.
result Consensus visualization effectively identifies shared low-dimensional data structure.
Survey reviews explainability in AI for healthcare, emphasizing trust and transparency.
problem Lack of transparency hinders AI adoption in healthcare.
method Comprehensive literature review to guide explainable AI design.
result Quantitative evaluation metrics are needed for some explainability properties.
AI helps HEP measure uncertainties, but needs better interpretation.
problem Uncertainty interpretation in AI for HEP measurements.
method Discussing existing AI methods for inference, simulation, and control/decision-making.
result Need for trustworthy AI UQ methods for widespread usage.
Develops a new approach for algorithmic recourse in AI systems.
problem Tackles the problem of providing recommendations for reversing negative AI decisions.
method Introduces a causal framework that models recourse as a process over pre- and post-intervention outcomes, allowing for partial stability and resampling of latent variables.
result Demonstrates the value of the proposed methods on real and semi-synthetic datasets.
Study improves summarization reliability in risky scenarios.
problem Reliability of automatic summarization in high-risk contexts.
method Conditional generation with Bayesian inference and entropy regularization.
result Significant improvement in robustness and reliability of summarization.
Conformal prediction improves signal detection accuracy in railway images.
problem Improving the reliability of machine learning models for railway signal detection.
method Applying conformal prediction to a novel dataset of train operator perspective images.
result The approach enhances the reliability of machine learning models for detecting railway signals.
New holistic approach measures sample-level adversarial vulnerability for trustworthy systems.
problem Inherent bias in adversarial attacks across subgroups.
method Combining high-frequency feature reliance and sample-distance to decision boundary.
result Holistic approach improves adversarial vulnerability estimation and system trustworthiness.
Paper quarantines unreliable Yelp users by detecting review spam.
problem Unreliable and spamming users deceive Yelp's users.
method Used RSD and spam detection techniques on key features.
result More than 80% of Yelp's accounts are unreliable, and highly-rated businesses are often spammed.
This work improves neural network trustworthiness through uncertainty estimation.
problem Overconfident neural networks lead to poor performance under distribution shifts.
method Develops a general uncertainty framework for neural networks, including classification with rejection.
result Improves model trustworthiness and robustness in decision-making tasks.
Research tackles distribution shift issues in ML to improve AI reliability.
problem Distribution shift limits ML reliability and trustworthiness.
method Study three distribution shifts (perturbation, domain, modality) and investigate robustness, explainability, adaptability.
result Proposes effective solutions and fundamental insights for enhancing ML robustness, adaptability, and safety.
In many contexts, it can be useful for domain experts to understand to what extent predictions made by a machine learning model can be trusted. In particular, estimates of trustworthiness can be useful for fraud analysts who process machine learning-generated alerts of fraudulent transactions. In this work, we present …
New framework makes ML methods compliant with regulations.
problem Ensuring ML methods meet regulatory standards.
method InfoGram and Admissible Machine Learning framework.
result Redesigns ML methods for regulatory compliance.
A new framework promotes trustworthy user-generated datasets by ensuring no user benefits from misreporting.
problem Incentivizing data misreporting in user-generated datasets.
method Proposes Licchavi, a global and personalized learning framework with provable strategyproofness guarantees.
result Proves that no user can gain much by replying to Licchavi's queries with deviated answers.
Deviation-based learning improves recommender systems by abstaining from recommending choices users might follow.
problem Recommender systems learn from user choices but can stall if users blindly follow recommendations.
method The recommender learns user knowledge by observing choices, abstaining from recommending a choice when multiple alternatives produce similar payoffs.
result Learning rate and social welfare improve when the recommender abstains from recommending certain choices.
New method improves OOD detection without sacrificing generalization.
problem Improving OOD detection models that also generalize well.
method Decouples uncertainty learning from Bayesian perspective.
result Achieves state-of-the-art OOD detection performance.
SMILE improves explainability of machine learning models.
problem Difficulty in understanding and trusting the conclusions of black-box machine learning models.
method Statistical Model-agnostic Interpretability with Local Explanations (SMILE).
result SMILE makes machine learning models more interpretable.
SNAP improves robust computation by emphasizing trustworthy items and downweighting outliers.
problem Improving robustness in computation, especially in high-dimensional settings.
method SNAP assigns weights based on mutual agreement, suppressing outlier contributions.
result SNAP ensures outliers contribute negligibly to computations, even in high-dimensional settings.
CEILS generates feasible counterfactual explanations by considering causal impacts.
problem Current counterfactual explanations lack feasibility and causal impact consideration.
method CEILS integrates causal reasoning into existing counterfactuals generation algorithms.
result CEILS provides feasible recommendations to achieve desired outcomes.