The ride-hailing market often experiences significant fluctuations in traffic demand, resulting in supply-demand imbalances. In this regard, the dynamic subsidy strategy is frequently employed by ride-hailing platforms to incentivize drivers to relocate to zones with high demand. However, determining the appropriate amount of subsidy at the appropriate time remains challenging. First, traffic demand exhibits high non-stationarity, characterized by multi-context patterns with time-varying statistical features. Second, high-dimensional state/action spaces contain multiple spatiotemporal dimensions and context patterns. Third, decision-making should satisfy real-time requirements. To address the above challenges, we first construct a Non-Stationary Markov Decision Process (NSMDP) based on the assumption of ride-hailing service systems dynamics. Then, we develop a solution framework for the NSMDP. A change point detection method based on feature-enhanced LSTM within the framework can identify the changepoints and time-varying context patterns of stochastic demand. Moreover, the framework also includes a deterministic policy deep reinforcement learning algorithm to optimize. Finally, through simulated experiments with real-world historical data, we demonstrate the effectiveness of the proposed approach. It performs well in improving the platform’s profits and alleviating supply-demand imbalances under the dynamic subsidy strategy. The results also prove that a scientific dynamic subsidy strategy is particularly effective in the high-demand context pattern with more drastic fluctuations. Additionally, the profitability of dynamic subsidy strategy will increase with the increase of the non-stationary level.

In the new normal of the supply chain, shortage risk is increasingly becoming a primary concern for operations management. While increasing order quantities can help alleviate shortage costs, surplus order quantity concurrently raises the risk of overstocking. Shortage and holding risk are affected by both demand and supply uncertainty, and the distributions of the random variables remain unknown. Consequently, solving the stochastic optimization problem analytically proves to be challenging. This study aims to exploit the data-driven stochastic optimization method and to analyze the optimal ordering decisions under different demand and supply scenarios. Numerical results indicate that the linear machine learning (LML) method incorporating data features outperforms the sample average approximation (SAA) method in terms of optimization performance. Incorporating $ l_2 $ regularization into the LML method effectively mitigates overfitting and produces similar results under different scenarios. In addition, the optimal order quantity may decrease as the supply variance increases when the unit shortage cost is relatively low, while the optimal order quantity may increase as the supply variance increases when the unit shortage cost is high. The opposite is true in the context of demand research. Finally, smaller supply and demand variances, together with larger supply means, lead to lower firm costs. The effect of the demand mean on the firm's costs is negligible. At the same time, the effect of the supply mean on the firm's costs is more pronounced than that of the supply variance.