A Hybrid Artificial Bee Colony Optimization Algorithm for Regional Economic Trend Prediction Based on Multi-Agent Systems and Induced Ordered Weighted Averaging Operator

With the acceleration of globalization and informatization, regional economic forecasting becomes more and more complicated and important. Traditional economic trend forecasting methods are often faced with the problems of high data complexity, difficulty in model selection and low accuracy of results. Therefore, how to effectively predict regional economic trends, especially in dynamic and changeable environments, has become the focus of research in economics, computer science and artificial intelligence. In this paper, a hybrid artificial bee colony optimization algorithm (Hybrid ABCO) based on the multi-agent system (MAS) and induced ordered weighted average (IOWA) operators is proposed, aiming at more accurate prediction of regional economic trends by integrating these technologies. First, we use MAS to simulate the interaction and economic decision-making process of different economic actors (such as enterprises, consumers, and governments), so as to reflect the complexity and dynamics of regional economy more truly. MAS can provide multi-dimensional information input for regional economy and provide diversified initial conditions for optimization algorithm by simulating the interaction of various economies in a region. Second, the IOWA operator is introduced into the model to process all kinds of information output by the MAS, and carry out reasonable weighting according to its weight, so as to improve the accuracy of regional economic prediction. IOWA operator can be dynamically adjusted according to the weight of historical data to further improve the flexibility and adaptability of the model. In the optimization process, we use Hybrid ABCO algorithm, which combines the global search capability and local optimization strategy of traditional ABC algorithm to improve the accuracy of regional economic forecast. The artificial bee colony optimization algorithm simulates the foraging process of bees, searches for the optimal solution through multiple iterations, and avoids falling into the local optimum by using the global search ability, which ensures the performance of the algorithm on large-scale datasets. The experimental results show that the method based on Hybrid ABCO is more accurate and robust than traditional forecasting methods such as support vector machine (SVM) and artificial neural network (ANN) in predicting regional economic trends. By introducing MAS and induced weighted average operator, the proposed method not only improves the prediction accuracy of the model, but also shows strong adaptability when dealing with complex regional economic data. The research in this paper provides a new way of thinking and method for regional economic forecasting, especially in the face of large-scale, multi-dimensional and dynamic data environment, it can effectively improve the performance of the forecasting system. In the future, the hybrid optimization algorithm proposed in this paper can be further applied to other fields, such as market trend prediction, socio-economic development planning and environmental change prediction.