Malicious software detection and model construction method and device, equipment and medium

The invention discloses a malicious software detection and model construction method and device, equipment and a medium, and is applied to the field of information security. The model construction method comprises the steps that a supervised learning model and a self-supervised learning model are constructed; calculating the labeled software sample through a supervised learning model to obtain cross entropy loss of a supervised side; calculating the unlabeled software sample through the self-supervised learning model by using comparative learning to obtain comparative learning loss; and carrying out fusion and joint training on the cross entropy loss and the contrast learning loss to obtain a malicious software detection model. According to the method, on the basis of supervised malicious software detection, a comparative learning framework in self-supervised learning is introduced to learn a large amount of label-free data in a production environment, so that the model can adapt to a rapidly changing malicio