Site flow optimization

1. A computer-implemented method comprising:

for each of a plurality of possible presented actions corresponding to a particular flow of actions in a computer network, and for each of one or more possible performed actions for each possible presented action, determining a likelihood that a user will perform the possible performed action;

weighting each of the determined likelihoods by applying a weight assigned to a corresponding possible presented action;

identifying an optimum presented action by determining a presented action having a weighted maximum determined likelihood, based on the weighted determined likelihood, wherein the identifying an optimum presented action includes utilizing a machine learning model having one or more user covariates, one or more performed action covariates, one or more contextual flow covariates, and interactions between at least two of the one or more user covariates, one or more performed action covariates, and one or more contextual flow covariates as input, the user covariates including information specific to the user, the one or more performed action covariates including information specific to each of the possible performed actions, the contextual flow covariates including information specific to the particular flow of actions; and

presenting the optimum presented action to the user in response to the particular flow of actions in the computer network.

2. The method of claim 1 , wherein the possible presented actions include web pages to which to redirect the user.

3. The method of claim 2 , wherein the web pages are located in a social network site to which the user is a member.

4. The method of claim 1 , wherein the one or more user covariates include demographic information about the user.

5. The method of claim 1 , wherein the one or more user covariates include profile information about the user.

6. The method of claim 1 , wherein the one or more user covariates include prior user actions.

7. A computer-implemented system comprising:

a processor;

a response prediction system executable by the processor and configured to:

for each of a plurality of possible presented actions corresponding to a particular flow of actions in a computer network, and for each of one or more possible performed actions for each possible presented action, determine a likelihood that a user will perform the possible performed action;

weight each of the determined likelihoods by applying a weight assigned to a corresponding possible presented action;

identify an optimum presented action by determining a presented action having a weighted maximum determined likelihood, based on the weighted determined likelihood, wherein the identifying an optimum presented action includes utilizing a machine learning model having one or more user covariates, one or more performed action covariates, one or more contextual flow covariates, and interactions between at least two of the one or more user covariates, one or more performed action covariates, and one or more contextual flow covariates as input, the user covariates including information specific to the user, the one or more performed action covariates including information specific to each of the possible performed actions, the contextual flow covariates including information specific to the particular flow of actions; and

an online social network system executable by the processor and configured to:

present the optimum presented action to the user in response to the particular flow of actions in the computer network.

8. The system of claim 7 , wherein the determining a likelihood utilizes a machine learning model taking input about the user from the online social network system.

9. The system of claim 7 , wherein the determining a likelihood utilizes a machine learning model taking outside input about the user from outside the online social network system, the outside input containing information not known to the online social network system.

10. A non-transitory machine-readable storage medium having instruction data to cause a machine to perform the following operations:

for each of a plurality of possible presented actions corresponding to a particular flow of actions in a computer network, and for each of one or more possible performed actions for each possible presented action, determining a likelihood that a user will perform the possible performed action;

weighting each of the determined likelihoods by applying a weight assigned to a corresponding possible presented action;

identifying an optimum presented action by determining a presented action having a weighted maximum determined likelihood, based on the weighted determined likelihood, wherein the identifying an optimum presented action includes utilizing a machine learning model having one or more user covariates, one or more performed action covariates, one or more contextual flow covariates, and interactions between at least two of the one or more user covariates, one or more performed action covariates, and one or more contextual flow covariates as input, the user covariates including information specific to the user, the one or more performed action covariates including information specific to each of the possible performed actions, the contextual flow covariates including information specific to the particular flow of actions; and

presenting the optimum presented action to the user in response to the particular flow of actions in the computer network.

11. The non-transitory machine-readable storage medium of claim 10 , wherein the possible presented actions include web pages to which to redirect the user.

12. The non-transitory machine-readable storage medium of claim 11 wherein the web pages are located in a social network site to which the user is a member.

13. The non-transitory machine-readable storage medium of claim 10 , wherein the one or more user covariates include demographic information about the user.

14. The non-transitory machine-readable storage medium of claim 10 , wherein the one or more user covariates include profile information about the user.