08-01-2014, 03:04 PM
The convergence of budding reckoning Soft computing technologies
Abstract:
Soft Computing methodologies represents the amalgamation of promising problem-solving technologies such as Neural Networks (NNs), Fuzzy Logic (FL), Probabilistic Reasoning (PR) and Genetic Algorithms (GAs) . We present a collection of tools and methods that can be used to perform diagnostics, estimation, and control. These tools are an enormous match for real-world applications that are characterized by imprecise, uncertain data and incomplete domain knowledge. Each of these technologies offer us with complementary reasoning and searching methods to solve complex problems. In this paper after a brief description of each of these technologies, we will analyze some of their most useful combinations, such as the use of FL to control GAs and NNs parameters; the application of GAs to evolve NNs or to tune FL controllers; and the implementation of FL controllers as NNs tuned by back propagation-type algorithms .We summarize the recompense of SC techniques and in particular the synergy derived from the use of hybrid SC systems.
Abstract:
Soft Computing methodologies represents the amalgamation of promising problem-solving technologies such as Neural Networks (NNs), Fuzzy Logic (FL), Probabilistic Reasoning (PR) and Genetic Algorithms (GAs) . We present a collection of tools and methods that can be used to perform diagnostics, estimation, and control. These tools are an enormous match for real-world applications that are characterized by imprecise, uncertain data and incomplete domain knowledge. Each of these technologies offer us with complementary reasoning and searching methods to solve complex problems. In this paper after a brief description of each of these technologies, we will analyze some of their most useful combinations, such as the use of FL to control GAs and NNs parameters; the application of GAs to evolve NNs or to tune FL controllers; and the implementation of FL controllers as NNs tuned by back propagation-type algorithms .We summarize the recompense of SC techniques and in particular the synergy derived from the use of hybrid SC systems.