Both artificial intelligence and additive manufacturing are excellent and revolutionary technologies. The main aim of this research paper is to predict surface roughness in additively manufactured processes in Poly Lactic Acid+ polymer material through different machine learning algorithms like support vector machine, linear regression, and two ensemble learning techniques Xtremegradient boosting and randomforest regressor. All machine learning is trained and tested. Predictive model of surface roughness is developed and machine learning behaviour is analyzed. Taguchi's Design of the Experiment was used to make L25 orthogonal array sample datasets. The machine model works on five parameters that influence layer geometries: layer height, infill density, printing speed, and nozzle temperature with a 0-degree raster angle. By applying all machine learning algorithms, random forest regression is best model, which gives 94.85% accurate results in datasets with minimum mean squared error of approximate 0.1255 and maximum r2 score of approximate. 0.9685.