This paper gives a brief presentation of the influence of different information structures on portfolio selection and some simple examples to demonstrate their basic algorithms. The first is the familiar mean_variance model. The model does not need more hypotheses on the motion of the asset price along the full continuous time axis,but sometimes many unanticipated computation problems may be involved because of the difficulty in computing the inverse of some matrices with high dimension. The second is the stochastic optimal control model. It is to solve a stochastic optimal control problem,which is often related to Hamilton_Jacobi_Bellman eqution.It results in solving partial differential equations of Riccati matrix differential equations. Their closed_form solutions are usually very hard to obtain. The last is the differential game model. The existence,uniqueness,and the computation for its equilibrium solution are all very difficult mathematical problems.