The Knowledge-Based Technology Applications Center (KBTAC) seminar series

Neural Networks are adaptive systems that allow users to model complex systems, classify patterns, filter data or control processes with little or no a-priori knowledge. Neural networks represent a branch of artificial intelligence technology. Fuzzy logic is a generalization of formalized mathematical logic that allows representation of uncertainty by providing a smooth transition from true to false, instead of a step change. It allows for a proposition to be both true and false, to different degrees at the same time. This allows representation of approximate reasoning concepts that occur frequently in everyday experience, such as ''somewhat true,'' or ''not very hot.'' To utilize these technologies, utility personnel need to acquire knowledge and skills in a number of areas: 1. basic principles of neural networks and fuzzy logic; 2. types of neural networks and fuzzy logic systems, their applications, and requirements for use; 3. considerations for designing neural networks and fuzzy logic systems; 4. software tools available for these technologies. Off-the-shelf software is available for a variety of hardware platforms to rapidly develop, tune and test neural networks and fuzzy logic systems. Familiarity with the problem to which a neural network or fuzzy logic system is applied allows the designer to develop an appropriate design. Therefore, it is desirable to teach neural network and fuzzy logic technology to utility experts in the domain of application