AskReference
DefinitionIntermediate

How is the F1 score calculated in the context of the LSTM RNN-based FDI detectors, and what do true positive, true negative, false positive, and false negative represent?

The F1 score is calculated using the harmonic mean of precision and recall: F1 = 2 × (Pr × Re) / (Pr + Re). Precision is TP/(TP+FP) and recall is TP/(TP+FN). In the LSTM RNN-based FDI detector, a true negative represents a correctly detected assault measure, a true positive represents an accurately discovered regular observation, a false positive represents a mistakenly observed assault, and a false negative represents a missed assault.

Key points

  • F1 = 2 × (Pr × Re) / (Pr + Re).
  • Pr = TP / (TP + FP).
  • Re = TP / (TP + FN).
  • TN = correctly detected assault measures; TP = accurately discovered regular observations; FP = mistakenly observed assaults; FN = missed assaults.
Source:AI and Machine Learning for Mechanical and Electrical Engineering ...· Data Science and ML Algorithms to Investigate Different Testing Scenarios for Various Anomalies in Driven Electric Motor· p. 69–71
Cover of AI and Machine Learning for Mechanical and Electrical Engineering ...

AI and Machine Learning for Mechanical and Electrical Engineering ...

Unknown

First edition · CRC Press

View this ebook