In uplink multi-user multi-input multi-output(MU-MIMO) network, user scheduling and transmit precoding are effective ways to manage the inter-user interference and antenna correlation. However, finding the optimal solution of user scheduling and transmit precoding matrix may need exhaustive search or iterative optimization, which may be impractical in realistic scenario, and suboptimal heuristic algorithms may lead to system sum rate performance degradation. In this paper, a deep deterministic policy gradient(DDPG) based method is proposed to select the appropriate scheduled user set and their transmit precoding matrices, then the system sum rate is improved. Moreover, instead of the channel coefficient matrix, the channel correlation measure matrix on behalf of the channel information is used as the state elements which can be further reduced by using the Hermitian feature of this matrix. Simulation results demonstrate that, compared with the benchmark algorithm, the proposed method can achieve a higher sum rate.