Designing the geometry of the force diagram to control the magnitude of the internal forces is a unique property of geometry-based structural design methods such as 3D graphic statics. For instance, subdividing the force diagram and its polyhedrons using various rules results in a variety of topologically different axially loaded structures for a given boundary condition with different load-bearing capacities. The solution space of all possible forms resulted from various subdivision techniques for a given boundary condition is vast. It is impossible to iterate through all possible forms to find solutions that can satisfy certain construction and loadbearing criteria within the time limit.Thus, in this research, we show how by using different predictive machine learning techniques, one can train a surrogate model to accelerate the structural performance assessment of various possible forms without the need to go through the slower process of geometric operations to iterate through a variety of solutions. Moreover, this process will result in more advanced sampling methods, where the machine learning models assist the designer in choosing different design strategies.