In this paper the multi-period mean semivariance portfolio problem is dealt with minimum transaction lots considering, transaction costs, borrowing constraints and threshold constraints. In this case the problem of finding a feasible solution is NP-complete. An optimal investment policy can be generated to help investors not only achieve an optimal return, but also have a good risk control. The multi-period portfolio selection is the mix integer dynamic optimization problem with path dependence. The forward dynamic programming method is designed to obtain the optimal portfolio strategy. Finally, the comparison analysis with borrowing risk-free assets and without risk-free assets in the portfolio selection is provided by a numerical example to illustrate the efficiency of the proposed approaches and the designed algorithm.