Infrared co-pyrolysis was employed to convert waste into valuable resource due to the increasing production of plastic waste. In this study, support vector machine (SVM) accurately predicted PVC and furfural residues thermal behavior. The results showed high accuracy (R2 > 0.95) after the SVM model was trained and validated. Furthermore, this research performed an in-depth characterized of the co-pyrolysis process using various analytical techniques, indicting TG-FTIR-GC/MS, XPS, simulated distillation, and GC/MS. As the temperatures and heating rates increased, the pyrolysis oil yield increased at first and then declined, reaching a maximum of 37.7 % at 600 °C and 10 °C/s. Notably, aromatic content was relatively high, comparing with other components, whereas the chlorinated content remained relatively low, all below 6.5 %, indicating the pyrolysis oil quality had effectively improved. According to the results of XPS, there was a significant 64.11 % increase in inorganic chlorination, which suggested the effective reduction in gaseous chlorine emission and achieving chlorine immobilization.