The invention belongs to the technical field of the maintenance and monitoring of an electric power device, and relates to an installation and debugging method of Beidou generation-II navigation system electric iron tower deformation monitoring device based on machine learning. The method comprises the steps: firstly performing one-dimensional linear regression analysis and least square method verification for iron tower acquisition data, carrying out signal-noise separation for debugging data by adopting a wavelet transformation method, establishing an auto-regressive and moving average (ARMA) model for the debugging data by adopting a time sequence analysis method, and finally evaluating, by experience, the prediction precision by adopting an evaluation assuming method, and analyzing and determining an optimum quantity, an optimum position and an optimum angle of a sensor.