Application of ACO in financial statement quality evaluation: From the perspective of corporate governance

This paper proposes an improved Ant Colony Optimization (ACO) algorithm to enhance the efficiency and accuracy of financial statement quality evaluation. Traditional ACO suffers from slow convergence and the risk of falling into local optima, particularly in multi-objective optimization problems involving complex financial data. To address these limitations, the proposed method incorporates Principal Component Analysis (PCA) for feature extraction, reducing the impact of redundant data and providing clearer input for optimization. A dynamically adjusted pheromone update strategy and a local search mechanism are introduced to strengthen the algorithm’s global search capability and prevent premature convergence. A multi-objective optimization model is developed, incorporating key financial indicators such as profitability, debt-paying ability, and management efficiency, with differentiated weights assigned to each objective based on its relevance in corporate governance. The algorithm’s performance is optimized through cross-validation and parallel computing, accelerating execution speed while minimizing time overhead in large-scale data processing. Experimental results demonstrate that the improved ACO algorithm achieves stable performance, converging close to the optimal solution within 40 iterations with an objective function value of approximately 0.9. Additionally, the algorithm yields a lower Mean Squared Error (MSE) of 0.285 across 10 different discount scenarios, indicating superior quality in financial statement evaluation. These findings validate the effectiveness of the proposed method in enhancing financial statement quality assessment.