JNCentenero/Multi-Fidelity-Autoencoder-UQ
PyTorch framework for multi-fidelity learning and uncertainty quantification. Combines a two-phase Multi-Fidelity Autoencoder with Multi-Split Conformal Prediction to transfer knowledge from abundant low-fidelity data to sparse high-fidelity data and provide reliable prediction intervals.