Luben M. C. Cabezas
@kuben45.bsky.social
Phd candidate in Statistics - (UFSCar/ICMC).
Mainly interested in Statistical ML, UQ in ML, Non-parametric inference, and foundations of Stats.
My website: https://monoxido45.github.io/
Mainly interested in Statistical ML, UQ in ML, Non-parametric inference, and foundations of Stats.
My website: https://monoxido45.github.io/
5/
EPICSCORE is accurate, flexible, and broadly applicable 💥
📄 Paper: arxiv.org/abs/2502.06995
#AI #ML #UncertaintyQuantification #ConformalPrediction #BayesianMethods
EPICSCORE is accurate, flexible, and broadly applicable 💥
📄 Paper: arxiv.org/abs/2502.06995
#AI #ML #UncertaintyQuantification #ConformalPrediction #BayesianMethods
Epistemic Uncertainty in Conformal Scores: A Unified Approach
Conformal prediction methods create prediction bands with distribution-free guarantees but do not explicitly capture epistemic uncertainty, which can lead to overconfident predictions in data-sparse r...
arxiv.org
May 7, 2025 at 3:23 PM
5/
EPICSCORE is accurate, flexible, and broadly applicable 💥
📄 Paper: arxiv.org/abs/2502.06995
#AI #ML #UncertaintyQuantification #ConformalPrediction #BayesianMethods
EPICSCORE is accurate, flexible, and broadly applicable 💥
📄 Paper: arxiv.org/abs/2502.06995
#AI #ML #UncertaintyQuantification #ConformalPrediction #BayesianMethods
4/
Strong results across tasks 📈
EPICSCORE adapts well to diverse settings—from regression to image classification—while improving uncertainty estimates 🔍✅
Strong results across tasks 📈
EPICSCORE adapts well to diverse settings—from regression to image classification—while improving uncertainty estimates 🔍✅
May 7, 2025 at 3:23 PM
4/
Strong results across tasks 📈
EPICSCORE adapts well to diverse settings—from regression to image classification—while improving uncertainty estimates 🔍✅
Strong results across tasks 📈
EPICSCORE adapts well to diverse settings—from regression to image classification—while improving uncertainty estimates 🔍✅
3/
We use Bayesian models—BART, GPs, MC Dropout— to adapt the interval width depending on data availability.
📊 More data → tighter intervals
🌌 Less data → wider intervals (epistemic uncertainty!)
We use Bayesian models—BART, GPs, MC Dropout— to adapt the interval width depending on data availability.
📊 More data → tighter intervals
🌌 Less data → wider intervals (epistemic uncertainty!)
May 7, 2025 at 3:23 PM
3/
We use Bayesian models—BART, GPs, MC Dropout— to adapt the interval width depending on data availability.
📊 More data → tighter intervals
🌌 Less data → wider intervals (epistemic uncertainty!)
We use Bayesian models—BART, GPs, MC Dropout— to adapt the interval width depending on data availability.
📊 More data → tighter intervals
🌌 Less data → wider intervals (epistemic uncertainty!)
2/
We built EPICSCORE:
✅ Works with any conformal score
✅ Adds Bayesian modeling of epistemic uncertainty
✅ Keeps all coverage guarantees
✅ Achieves asymptotic conditional coverage
We built EPICSCORE:
✅ Works with any conformal score
✅ Adds Bayesian modeling of epistemic uncertainty
✅ Keeps all coverage guarantees
✅ Achieves asymptotic conditional coverage
May 7, 2025 at 3:23 PM
2/
We built EPICSCORE:
✅ Works with any conformal score
✅ Adds Bayesian modeling of epistemic uncertainty
✅ Keeps all coverage guarantees
✅ Achieves asymptotic conditional coverage
We built EPICSCORE:
✅ Works with any conformal score
✅ Adds Bayesian modeling of epistemic uncertainty
✅ Keeps all coverage guarantees
✅ Achieves asymptotic conditional coverage
1/
Conformal prediction is great for distribution-free coverage ✅
But it misses epistemic uncertainty—when data is sparse, and the model just doesn’t know 🤷
Most fixes are task-specific (regression, quantile) and hard to generalize.
Conformal prediction is great for distribution-free coverage ✅
But it misses epistemic uncertainty—when data is sparse, and the model just doesn’t know 🤷
Most fixes are task-specific (regression, quantile) and hard to generalize.
May 7, 2025 at 3:23 PM
1/
Conformal prediction is great for distribution-free coverage ✅
But it misses epistemic uncertainty—when data is sparse, and the model just doesn’t know 🤷
Most fixes are task-specific (regression, quantile) and hard to generalize.
Conformal prediction is great for distribution-free coverage ✅
But it misses epistemic uncertainty—when data is sparse, and the model just doesn’t know 🤷
Most fixes are task-specific (regression, quantile) and hard to generalize.