Intelligent Systems: Architectures and Perspectives

Title: Intelligent Systems: Architectures and Perspectives

Research Question: How can different learning and adaptation techniques be integrated to create more effective and efficient intelligent systems?

Methodology: The study investigates the design of intelligent systems by combining various techniques such as Neural Networks (NN), Fuzzy Inference Systems (FIS), Probabilistic Reasoning (PR), and Evolutionary Computation (EC). It presents different architectures for integrating these techniques, including NN-FIS, EC-FIS, EC-NN, FIS-PR, and NN-FIS-EC systems.

Results: The study found that hybrid systems can achieve better performance by combining the strengths of different techniques. For example, NN can store and generalize knowledge, while FIS can handle symbolic reasoning and uncertainty. EC can optimize the system and handle complex problems, while PR can provide probabilistic reasoning.

Implications: The research suggests that a common framework for designing intelligent systems can lead to better comparison and evaluation of different hybrid architectures. It also highlights the importance of focusing on the integration and interaction of different techniques rather than merging them to create new methods. This approach can help in developing more effective and efficient intelligent systems.

Link to Article: https://arxiv.org/abs/0405009v1 Authors: arXiv ID: 0405009v1