Porous metals are increasingly important in technology. Due to their tunable mechanical properties, they are promising candidates in various emerging applications such as metallic scaffolds for load-bearing bones, lightweight structures for transport technologies, electrodes for electrochemical energy storage devices and more. This project aims at developing a computational model to establish a quantitative understanding of the relations between the enormous variety of possible morphologies of porous structures and their mechanical properties. The machine-learning based model will be employed to identify various prototype structures of new morphologies via implementation of hierarchical screening and a material genome approach. The optimal and/or statistically relevant structures will be 3D printed and tested mechanically in experiments, with the results contributing to both tuning and validation of the computational designs.