Neural network application and expert
system
Many
difficulties in business world are vitally about endeavoring to forecast the
prospect of distinct outcomes. The real world is so convoluted, with so
numerous inter-related variables that predicting these conclusions are often
very tough. Conventional computing has a limited function to play in assisting
with these jobs but neural networks offer an alternate answer. By feeding in the
different components that offer conclusion over time can investigate preceding
trends and patterns to forecast the future.
G. R. Pugh
& business does consulting to predict the charges of bonds of public
utilities. The business utilized neural network to help forecast the following
years business bond prices and rankings of over 100 public utility companies.
The network they utilized contrasted very favorably to accepted mathematical
investigation. The network was able to forecast a utility rating (A, B, C) with
95% correctness, accepted mathematical investigation was only productive 85% of
the time. The only adversities came across by the mesh were associated with
companies experiencing particularly unusual problems that were not incorporated
into the systems inputs.
Medical
diagnosis is an example of expert system the knowledge groundwork would comprise
health data, the symptoms of the patient would be utilized as the query, and
the recommendations would be a identify of the patient’s sickness The software
used in medical apparatus controls and interacts exactly with the physical
natural environment, it is subject to firm asset constraints, it must function
in real-time, it desires to persuade many safety requirements, as well as other
obligations such as reliability and availability.
No comments:
Post a Comment