Skip to content
RT
Skip to main content

Research Q&A

66 main questions · 539+ sub-questions · Linked to thesis evidence

Problem → Framework → Results → Policy
AI / Machine Learning
Optimization
Forecasting
Energy
Statistics
Policy
Methodology
📖

Select a concept from the left to view its definition.

605 questionsThesis →

↑↓ Navigate, Esc Clear filters

D8Layer 5 / 10hard~4 min read

What is Monte Carlo dropout for uncertainty?

Stochastic forward passes at test; dropout remains active. 100 passes → prediction intervals.

Key Takeaway
PICP=95.4%, MPIW=11.8 GW.
Detailed Explanation
MC dropout: keep dropout active at inference; sample ; with T=100 passes. Captures epistemic (model) uncertainty. Prediction interval: where μ, σ from empirical distribution. PICP=95.4%, MPIW=11.8 GW. Bayesian approximation without full BNN.
Why It Matters
Research significance: UQ method.
Energy policy relevance: Risk-aware.
Methodological contribution: D1.
Research Graph — Related Questions
Research Q&A — Appendix | Renewable Transition