Electroactive polyvinylidene fluoride (PVDF) with predominantly the β-phase is now challenging the fabricating of PVDF toward energy storage applications. Here, the comprehensive effect of BaTiO3 nano-particles and mechanical stretching on the improvement of the β-phase of PVDF was investigated. In situ synchrotron wide-angle X-ray diffraction (WAXD) and small-angle X-ray scattering (SAXS) measurements were performed to investigate this behavior. Consequently, the transformation rate of β-crystal for both pure PVDF (BT0) and PVDF/BaTiO3 (90/10) nanocomposites (BT1) decreases as the stretching temperature increases, implying that the high temperature is unfavorable to the formation of β-crystal phase. A synergistic enhancement of the β-phase of nano-particle additives and stretching was discovered. It was surprisingly observed that α-phase of PVDF would completely transform into β-crystal in BT1 sample at 100 °C during stretching. The mechanism of the synergistic effect of BaTiO3 nano-particles and mechanical stretching was proposed. Moreover, machine learning was implemented to predict the fraction of β-crystal phase (F(β)) of the PVDF/BaTiO3 composites under various uni-axial stretching conditions by Python 3.8. The results show that the machine learning technique can rapidly and efficiently discover the ideal value of F(β) and the optimal multivariate coupling conditions.