Back to feed
arXiv cs.LG
arXiv cs.LG
7/7/2026
Weighted Conformal Prediction for Lab-to-Track Thermal Transfer in EV Motorsport Powertrains

Weighted Conformal Prediction for Lab-to-Track Thermal Transfer in EV Motorsport Powertrains

Short summary

Researchers develop weighted conformal prediction to solve the 'lab-to-track' gap in EV powertrains, where models trained on controlled lab cycles fail under real-world loads. Using density-ratio weighting with ensemble prediction intervals, they improve coverage from 70.13% to 72.42% under covariate shift, validated on CALCE lithium-ion cell data and Formula 1 telemetry. The approach is partial—conformal domain adaptation is promising but incomplete.

  • •Weighted conformal prediction bridges EV thermal modeling gap between lab testing and real-world driving
  • •Modest improvement: 70% to 72% coverage under domain shift using ensemble density-ratio methods
  • •Honest assessment: method is promising but only partially solves the transfer learning problem

Generated with AI, which can make mistakes.

#research-breakthrough#ai-tools#industry-adoption
Read full article at arXiv cs.LG

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

A Quiet Failure in Calibrated Virtual Screening: Marginal Conformal Prediction Under-Covers the Minority Class, and a Class-Conditional Fix Recovers It
arXiv cs.LG
BattVAE-GP: Generative Modeling of Long-Horizon Battery Degradation with Uncertainty Quantification
arXiv cs.LG
Evaluating Time Series Foundation Models for Electricity Price Forecasting: Contamination Risk, Distributional Shifts, and Covariate Dependence
arXiv cs.LG
Conformal Orbit-Valid Trust Horizons for Equivariant World Models
arXiv cs.LG