The skin is an important organ for human beings to perceive information from the outside world. Human beings use the sense of touch produced by their skin to judge the environmental and react accordingly. The Polyvinylidene fluoride(PVDF) piezoelectric fiber with liquid core and electronic skin flexible tactile sensor were designed and fabricated with the PVDF as the main material. In order to test the tactile sensing performance of the electronic skin, a in the experiment, a thrust meter is used to apply pressure, which makes the piezoelectric fiber deform, as a result, the charge changes. The data are trained by the machine learning error back propagation (BP) neural network algorithm and the Classification Learner, so that the size and position of pressure can be determined by the signal size of the six fibers. The regression coefficient(R) value in the regression diagram of BP neural network is 0.87, and the area under the curve (AUC) value in the receiver operating characteristic (ROC) curve of the Classification Learner is 1. The results show that the electronic skin has high sensitivity and accurate force capture.