Distributed logistic regression using expectation propagation.

Clients and server are connected through socket for performing distributed logistic regression.

We design a distributed logistic regression model using expectation propagation, where clients and server are connected through sockets. The server can handle the users' authentication with predefined user name and password. The proposed framework provides a high level guarantee for protecting sensitive information, since the information exchanged between the server and the client is the posterior distribution of coefficients.