<p>Dental caries, one of the most common oral diseases, is mainly induced by multiple cariogenic bacteria in the oral microenvironment, so it is important to construct a method that can identify oral multiply cariogenic bacteria. Herein, a machine learning assisted colorimetric sensing array has been constructed by using polyethylene glycolized graphyne nanozymes, which have been firstly prepared through click chemistry to enhance the peroxidase-like catalytic activity and stability. Taking advantage of different interactions between cariogenic bacteria and polyethylene glycolized graphyne nanozymes, the sensing array can well identify ten oral cariogenic bacteria and their binary and ternary complexes by linear discriminant analysis algorithm. In view of its high accuracy to analyze the spiked artificial saliva samples, the developed colorimetric sensing array owns a great potential for the diagnosis of dental caries.</p> Graphical Abstract <p></p>