An Approach to NILM using image-based features and transfer learning

In this contribution we investigate possibilities of using neural network classification methods for Non-Intrusive Load Monitoring (NILM) methods using V-I diagrams. For analysis we chose PLAID2018 dataset and transfer learning approach using deep neural networks. The results show the competitiveness of this approach against classical approaches as KNN or SVM. Effect of input feature vector dimensionality and additional feature enrichment is analysed.