Sorghum-legume intercropping has the potential to improve forage productivity, resource use efficiency and forage quality under irrigation in the Southern High Plains, USA. Crop production is conversion of solar radiation into biomass and solar radiation is wasted early in the season in most annual row cropping systems. An intercropping system improves radiation interception as well as utilization of water, nutrient and other resources. As a result, system efficiency is expected to be improved early in the season; however, later competition for resources including light reduces complementarity. Many crops try to adjust to low light by changing plant architecture, biomass partitioning and plant physiology to maintain productivity under shade. However, each species may differ in their response to shade. An ideal legume species for intercropping is the one that is least affected by shade while significantly contributing to forage quality and quantity. If it can improve productivity under low light it is even better. Support vector machine (SVM) is one of the statistical learning algorithms based on the statistical learning theory. SVM is rarely used in agronomic research for identifying and analyzing treatment effects. The main objective of the research was to study effects of shade on morphology, physiology and productivity of five diverse legumes (lablab, limabean, cowpea, pigeonpea and polebean) grown under shaded and unshaded conditions using SVM. A set of statistical relationships were developed and compared to evaluate the shade effects on morphology, physiology and productivity to identify suitable sorghum-legume intercropping system for forage production in the Southern High Plains. Preliminary analysis of the statistical results indicates that sorghum-lablab intercropping system followed by sorghum-limabean have the potential to improve forage productivity, resource use efficiency and forage quality under irrigation.