Abstract
This work aims to enhance the accuracy and efficiency of corporate strategic decision-making, particularly in rapidly changing and highly competitive market environments. Traditional strategic decision-making methods rely on managers' experiential judgment and exhibit limitations when handling complex data and high-frequency market fluctuations. To address this issue, this work proposes a hybrid optimization model combining transformer models and reinforcement learning algorithms, designed to optimize corporate strategic decision-making processes and improve competitiveness. First, relevant studies on strategic decision-making and corporate competitiveness are reviewed, clarifying the potential and advantages of artificial intelligence (AI) in decision support. Second, the hybrid model is developed and trained through steps including data collection and preprocessing, algorithm selection and model construction, as well as model training and validation. Finally, real-world data are applied to evaluate model performance across indicators such as training time, convergence speed, and prediction effectiveness. The results demonstrate that the hybrid model successfully converges within 150 iterations and exhibits substantial advantages over traditional algorithms, particularly in prediction accuracy for market share (92%), profit growth rate (91%), and customer satisfaction (89%). Implementing the model leads to notable improvements in corporate market position, brand influence, and technological innovation capabilities. The work shows that the hybrid model enhances the scientific rigor and accuracy of decision-making. Meanwhile, it strengthens corporate competitiveness and market responsiveness, highlighting the substantial potential of AI technologies in strategic management. This work provides enterprises with an efficient and reliable decision-support tool, facilitating the maintenance of competitive advantages in complex and dynamic market environments.