Fig. 6From: Fog computing application of cyber-physical models of IoT devices with symbolic approximation algorithmsBOSS model and BOSS VS: a Samples are being scanned with a sliding window, b multiple windowed subsequences are generated, c all of the subsequences are transformed into SFA words, d SFA words are summarized in the form of BOSS histogram (BOSS model), and e the BOSS histogram is vectorized through Term Frequency Inverse Document Frequency (TF-IDF) model, which finally results in TF-IDF vectors for training dataBack to article page