Integrating Complex Fuzzy Rough Yager Aggregation Operators with Neural Networks for Improved AI Predictive Models

When it comes to predictive modeling, choosing the most suitable neural network models is even more important in the age of expanded artificial intelligence. Neural networks as basic components of the modern AI systems themselves presuppose the proper selection of the framework regarding multiple criteria starting from the scalability and ending with computational performance and integration possibilities. This chapter meets this crucial challenge in two ways: first, by proposing a new MCDM framework for the selection of the neural network framework; and second, by developing the theoretical underpinnings of the complex fuzzy rough set theory. We propose four innovative aggregation operators: The proposed novel operators are complex fuzzy rough Yager power averaging (CFRYPA), complex fuzzy rough Yager power weighted averaging (CFRYPWA), complex fuzzy rough Yager power geometric (CFRYPG), and complex fuzzy rough Yager power weighted geometric (CFRYPWG). These operators expand conventional methodologies by including advanced fuzzy rough information processing functions which are useful in making enhanced decisions during situations involving uncertain and imprecise data. This is supported by the theoretical analysis of the properties of these operators and their mathematical proofs that describe their performance. As a proof of concept of the proposed framework, we provide a detailed case study of the selection of neural network frameworks for AI predictive models. The case study demonstrates how our MCDM technique can process fuzzy rough information to assess and rank the best neural network frameworks according to the criteria. In addition, we also perform a detailed comparison with the other methods to show the effectiveness and improvement of the proposed method. The combination of complex fuzzy rough Yager power operators with the selection of the 34 neural network framework is a major improvement in theoretical and practical aspects of AI system development. This chapter narrows the gap between theoretical mathematical models and practical AI applications, offering practitioners effective and reliable instruments to make proper decisions in the NN framework choice and, at the same time, developing the field of fuzzy mathematics and decision-making theory.