Market forecasting capability, as an important ability for enterprises to gain insight into changes in the external environment, directly affects the quality of strategic decisions. However, existing studies have not explored enough the mechanism of how market forecasting capability specifically affects the accuracy of strategic decisions, and lack a systematic quantitative analysis method. In this study, the ACO-MPSO association rule mining algorithm is designed by combining the ant colony optimization algorithm and particle swarm optimization algorithm and introducing the Metropolis mechanism. 1640 data of A-share listed companies in the manufacturing industry from 2016 to 2023 are selected, a decision table containing 25 conditional attributes is constructed, and the new algorithm is applied to mine the association rules of market forecasting ability and strategic decision precision. The results show that: when the enterprise market prediction ability is strong, 12 high strategic decision precision rules are mined, with an average accuracy of 90.4%; compared with the traditional Apriori algorithm, the ACO-MPSO algorithm completes the mining in only 15.84 seconds under the 45% support threshold, which is a 42% improvement in the efficiency; and the validation of the test set shows that the overall classification accuracy of the rules is 84.2%. Among them, the classification accuracy for high-precision samples reaches 93.16%. It is found that policy sensitivity, data accuracy, and resource endowment fulfillment are the key factors to promote high strategic decision-making accuracy, and the improvement of enterprise market forecasting ability can significantly enhance strategic decision-making accuracy.
