KPCA and STFT Non-intrusive Load Monitoring

Codes used to generate KPCA and STFT for Non-intrusive Load Monitoring

This work proposes an approach to extract features of electric current waveforms to identify residential appliances. The Short-Time Fourier Transform (STFT) and the kernel PCA technique were used to extract these features. Once the features have been defined, the classifiers k-Nearest Neighbor (kNN), Support Vector Machines (SVM), Linear Discriminant Analysis (LDA), Random Forest (RF) and Extreme Learning Machine (ELM) were used for appliance (or combination of appliances) identification.

PS: The ELM algorithm was extracted from http://www.ntu.edu.sg/eee/icis/cv/egbhuang.htm and adapted to this work