NILMPEds: A Performance Evaluation Dataset for Event Detection Algorithms in Non-Intrusive Load Monitoring

NILMPEds (NILM Performance Evaluation dataset), is a different type of NILM dataset, in a sense that it is aimed primarily at research reproducibility with respect to the development and performance evaluation of event detection algorithms. NILMPEds contains the results of 47950 event detection models when applied to four public event detection datasets. The different parameter configuration of each model and the ground-truth data are also available. Finally, this dataset also contains the performance evaluation of each model according to 31 performance metrics.