<p>Additive manufacturing offers higher precision, improved product quality, and time-efficient operations. However, selecting optimal process parameters to achieve desired mechanical properties in the final product often relies on trial-and-error methods, contributing to material wastage. Hence, this study focused on developing a bidirectional machine learning model that integrates the final mechanical properties of PLA + parts with key fused deposition modeling (FDM) process parameters. The novelty of this work lies in the model’s bidirectional predictive capability, which allows estimation of the optimal process parameters required to achieve desired mechanical properties and, conversely, prediction of mechanical properties based on given process parameters. For this study, polylactic acid plus (PLA +) was selected. A comprehensive factorial experimental design was implemented, focusing on four critical FDM process parameters: layer thickness, printing speed, infill density, and extrusion temperature. Each parameter was examined at four levels, with each combination replicated three times, yielding 768 samples. Following ASTM standards, these samples were tested for tensile strength, flexural strength, and longitudinal shrinkage. The model exhibited high predictive accuracy, with <i>R</i>² scores exceeding 0.97 and 0.99 for most parameters during the testing and training phases. Again, correlation analysis identified significant relationships between mechanical properties and input parameters. Infill density and tensile strength are positively correlated, whereas layer thickness is negatively correlated. The model’s prediction potential will enable manufacturers to make informed decisions on process parameters that will minimize material wastage and production costs and enhance sustainability.</p>