Cable wear in cable-driven parallel robot (CDPR) is a major concern regarding the safety of end-users for assistance applications or for preventive maintenance in industrial applications. To estimate this wear we propose to calibrate the Young modulus of the material of each cable, that is affected by the wear, using an approach based on the partial measurements of the platform pose at specific calibration poses. These measurements may be used to identify the Young modulus either with a deterministic algorithm or with a neural network with 2 variants. The deterministic method is more accurate and robust with respect to uncertainties but is computer intensive while the neural network is less accurate and robust but is almost real time.