Figure 3.

Flow chart of the proposed data-driven diagnosis methodology based on deep feature learning for fault diagnosis applied to the diagnosis and identification of bearing faults for different bearing technologies, such as metallic, hybrid and ceramic bearings, in electromechanical systems.