3D printers are used widely across various fields, including biomedical applications, electrical, electronics, architecture, textiles, art, and aerospace and aviation technologies. With the growing prevalence of 3D printers, optimizing print quality has become increasingly important. Before starting the production processes, mechanical and physical tests on samples are necessary to assess roughness, heat resistance, tensile strength, and elongation. Although these tests improve the print quality, they require printing new material samples each time. Constantly producing test samples leads to longer production times and material waste. This study analyses a dataset consisting of different print parameters and mechanical test results using machine learning methods to address this issue. The print quality parameters include layer height, wall thickness, infill density, infill pattern, nozzle temperature, bed temperature, print speed, material type, and fan speed. The parameters predicted by machine learning methods are roughness, tensile strength, and elongation. Due to the small dataset, Gaussian noise was added to generate an augmented dataset, and both the original and augmented datasets were evaluated. The gradient boosting algorithm produced good results for both datasets among the machine learning methods. This approach significantly reduces the need to print test objects, minimizing production time and material waste. It resulted in an improvement of nearly 30% in accuracy for roughness prediction, thereby reducing the need for test object production and making the 3D printing process more efficient and sustainable.