A Recommendation Model for College Career Entrepreneurship Projects Based on Deep Learning

Algorithm 1

Optimization algorithm based on the deep neural network and matrix decomposition.

Input: : user rating matrix

 : item attribute feature data

Output:

 : user implicit feature vector

 : implicit feature vector of items

Initialization:

 Regularization parameter:

 Maximum number of iterations: max_iter

 Dimension of implicit features: latent_dim

 Proof of implicit characteristics of randomly generated users and projects:

 Parameter set of the deep neural network:

 Calculate the loss value according to the formula:

For from 1 to max_iter do:

 For each in trainset

  Update

   Update

 End for

 Training DNN (deep neural network)

 Calculate the loss value according to the formula

 If

   Termination of iteration

End for