“I do not know”: Quantifying Uncertainty in Neural Network Based Approaches for Non-Intrusive Load Monitoring
This repository contains code for “I do not know”: Quantifying Uncertainty in Neural Network Based Approaches for Non-Intrusive Load Monitoring. This paper was accepted at BuildSys 2022.
Please use the following bib entry to cite the paper.
@inproceedings{10.1145/3563357.3564063,
author = {Bansal, Vibhuti and Khoiwal, Rohit and Shastri, Hetvi and Khandor, Haikoo and Batra, Nipun},
title = {"I Do Not Know": Quantifying Uncertainty in Neural Network Based Approaches for Non-Intrusive Load Monitoring},
year = {2022},
isbn = {9781450398909},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3563357.3564063},
doi = {10.1145/3563357.3564063},
pages = {79–88},
numpages = {10},
keywords = {calibration, neural networks, non-intrusive load monitoring, uncertainty, bayesian analysis},
location = {Boston, Massachusetts},
series = {BuildSys '22}
}
First, links to the notebooks for the figures
| Figure | Link |
|---|---|
| Figure 2 and 3 | Recalibration for Regression |
| Figure 4 | Dishwasher low MAE justification |
| Figure 5 | Boxplot for ECE values across appliances |
| Figure 6 | Fridge : Predicted power and reliability diagrams |
| Figure 7 | Dishwasher : Predicted power and reliability diagrams |
| Figure 8 | Microwave : Predicted power and reliability diagrams |
| Figure 9 | Fridge : Recalibration |
| Figure 10 | Dishwasher : Recalibration |
| Figure 11 | Microwave : Recalibration |
| Figure 12 and 13 | Poor Recalibration |
