“…HMM, a powerful pattern recognizer, classifies the events in a probabilistic manner based on fault signal waveform and characteristics [15]. The HMM ability to solve small sample, nonlinear, and high-dimensional pattern problems make this algorithm a powerful choice for application in power system disturbance classifications [16,17], partial discharge de-noising [18], accidents identification and decision making in power plants [19], modeling and forecasting electrical power markets [20][21][22], and power transformer fault diagnosis based on dissolved gas analysis [23]. In [24,25], HMMs were specially applied to power transformer differential protection.…”