In this experimental investigation, a study of water droplet dynamics has been performed on a semi-hierarchical textured cylindrical target. Three different textures, viz. cycloid, sawtooth, and triangular, with amplitude of 300 µm and time-period of 600 and 800 µm, were fabricated on a Nickel-Titanium alloy through the Wire Electrical Discharge Turning process. The contact angle of the textured samples improved drastically compared to the as-received samples and was related to the roughness factor (r). The spreading in the axial direction has been studied for the various textures, and it was observed that the textures significantly reduce the maximum spreading diameter compared to the smooth cylinders. Furthermore, the experimental investigation revealed splashing in the case of textured samples and deposition in the case of as-received samples at α ≥ 38.8, indicating r, along with diameter ratio, D* , plays a critical role in determining the splashing zone. Accordingly, a regime map has been plotted between r and α, and a critical curve in the form of hyperbolic function was observed. The average velocity of spreading and receding was calculated, and the spreading velocity was 3–5 times higher than the receding velocity. Finally, a model has been developed based on data-driven machine learning, which was in good agreement with the previous experimental investigations.