As the fiber reinforced polymer matrix composites gives good strength and can work in rigorous environmental conditions, nowadays more focus is given to study the behavior of these materials under different operating conditions. Due to the environmental concern focus on the natural fiber reinforced polymer matrix composite (NFRPC) is enhancing both in research and industrial sectors. Currently focus has been given to unify solid fillers with the NFRPC to improve its mechanical and tribo properties. Aligned to this, the present work discusses the effect of various weight fraction of fillers (Flyash, SiC and graphite) on the frictional behavior of natural fiber (cotton) polyester matrix composites. The specimen prepared with Hand lay-up process followed by compression molding. A plan of experiments, response surface technique, was used to obtain response in an organized way by varying load, speed and sliding distance. The test results reveal that different weight concentration of fillers has considerable result on the output. The frictional behavior of materials evaluated by general regression and artificial neural network (ANN). The validation experiment effects show the estimated friction by using ANN was more closer to experimental values compare to the regression models.