This directory contains an interactive notebook and a data file to inspect all results summarized in the following ICLR2021 submission:
Title: Uncertainty Prediction for Deep Sequential Regression Using Meta Models

Files: 
trimmed.pkl                    - a pickled dictionary with results data
ResultsVisualization.ipynb     - Python 3 notebook to browse and visualize the content of the above file
README                         - this README file

Requirements: Python 3, numpy, pickle, matplotlib
Compatibility: The data file was creaed using pickle version 4.0.

NOTE: Due to the upload limit of 100MB, the pickle file contains only the first 500 (of 1000) simulations for the SPE9PR dataset. The MITV dataset is full.  

