One of the main issues in Industry 4.0 is about data and ways to gain valuable information from them. Over the last few years, there has been an ongoing trend towards digitalization and data acquisition of polymer processing machineries. These developments brought a large amount of production data. In polymer processing, there are hundreds of influences that can affect product quality in many different ways. The main question is how to find these influences, which are the most important, which can be neglected, and which can be merged due to same origins. This paper addresses the use of methods to analyse experimental data. Experiments were performed on a highly digitalized and process-monitoring profile extrusion plant in and beyond the limits of standard processing. Influences of all process parameters on inline measured quality parameters, such as colour, gloss, profile weight, and the outer profile geometry, have been analysed. For analysis, we applied correlation matrices to detect linear relations and the machine learning algorithm random forest regression (RFR) to detect nonlinear relations. To prove the obtained correlations we observed the data via Scatter Plots and Line Charts. Further impact factors of RFR-models and correlation coefficients were analysed and proved. Advantages and disadvantages of those methods, as well as limitations in terms of usage, are shown.