In a gas turbine fault diagnostics, the removal of measurementnoise and data outliers prior to the fault analysis is very essential. Theconventional filtering methods, particularly the linear ones, are notsufficiently accurate, which might possibly lead to the loss of criticallyimportant features in the fault analysis process. Conversely, the recordedaccuracies obtained from the non-linear filters are promising. Recently, thefocus has been shifted to the artificial neural network (ANN) based nonlinearfilters due to their capability of providing a robust identity map between theinput and output data, which can be efficiently exploited in the process offault diagnosis. This paper aims to present combined auto-associative neuralnetwork (AANN) and K-nearest neighbor (KNN) based noise reduction andfault detection method for a gas turbine engine application. The performanceof the developed method has been evaluated using data obtained from amodel simulation. The test results revealed that the developed hybrid methodis more effective and reliable than the conventional methods for the faultdetection of the gas turbine engine with negligible false alarms and misseddetections.