Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Reliable Uncertainty Quantification in Machine Learning via Conformal Prediction
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
A doctoral consortium proposal outlining a research plan to extend conformal prediction to non-i.i.d. data, LLMs, improved predictive efficiency, and integration into training, with no reported experimental results.
Problem
Conformal prediction's guarantees rely on assumptions (i.i.d. data, pre-trained models) that are violated in real-world settings, and CP is underdeveloped for large generative models and for integration into model training.
Approach
Proposes a research agenda to adapt CP to non-i.i.d. settings (e.g., federated learning via similarity measures and importance weighting), extend it to LLMs, improve prediction set efficiency via non-conformity score analysis, and use CP as a training regularizer (conformal training).
Key results
- No quantitative results are reported; the paper states that code implementation on CIFAR-100 is underway as of September 30, 2024.
Limitations
- No experimental results; work is at proposal stage
- Non-i.i.d., LLM, and efficiency directions are planned but not yet validated
- Theoretical frameworks described as incomplete
Experimental setup
Datasets: CIFAR-100
Builds on: split conformal prediction conformal training
Closest papers by meaning
- Robust Uncertainty Quantification Using Conformalised Monte Carlo Prediction (AAAI 2024 · similarity 0.90)
- Conformal Prediction as Bayesian Quadrature (ICML 2025 · similarity 0.90)
- Verifiably Robust Conformal Prediction (NeurIPS 2024 · similarity 0.90)
- Generative Conformal Prediction with Optimized Coverage Allocation (NeurIPS 2026 · similarity 0.89)
- Conformal Prediction Beyond the Seen: A Missing Mass Perspective for Uncertainty Quantification in Generative Models (NeurIPS 2025 · similarity 0.89)
- An Information Theoretic Perspective on Conformal Prediction (NeurIPS 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.