Product recommendation method and device, equipment and medium

The invention provides a product recommendation method which can be used in the financial field or other fields. The method comprises the following steps: acquiring behavior data of M clients operating N products at each moment in T moments to obtain a behavior data set; based on the behavior data set, obtaining T original score matrixes corresponding to the T moments; based on the T original scoring matrixes, respectively extracting a customer feature vector sequence of each customer corresponding to T moments and a product feature vector sequence of each product corresponding to T moments; respectively processing the customer feature vector sequence and the product feature vector sequence by using two neural network models of the same type to correspondingly obtain a customer prediction score vector and a product prediction score vector; and on the basis of correlation analysis of the customer prediction score vector of each customer in the M customers and the product prediction score vectors of the N produ