Perhaps the largest-scale application of neural networks in business today is the optimal airline occupancy system from BehavIouristics in Silver Springs, Maryland. The program helps airlines to allocate seats by class most advantageously for each specific flight, taking into account complex combinations of various factors such as flight time, competitive conditions, passenger absences, etc. According to K. Stephens, Chairman of the Board of BehavIouristics, the program will be reviewed based on the company's available statistics.

Another area where intensive work is underway on the implementation of neural systems is the management of production processes. For example, A1 Ware in Cleveland has installed neural networks in chemical plants of one of the major industrial companies. With the help of these networks, which analyze the effect of different ratios of ingredients on the properties of the final product, the company's specialists created new types of products. GTE Laboratories in Waltham, Massachusetts, plans to install a network at a factory that produces fluorescent lamps. "We want to find out why the factory works well in some periods of time and poorly in others," said a representative of GTE, R. Sutton.

According to E. Rosenfeld, editor of the journal Intelligence, who monitors the state of affairs in the field under consideration, more than a hundred large companies are currently actively researching the possibility of using adaptive networks. The Defense Advanced Research Project Agency of the Ministry of Defense is also going to seriously engage in developments in the field of neural networks. However, as the agency's representative said, the annual budget allocations for these purposes in the next 8 years will be less than $400 million, despite the statement made by the agency's assistant Director J. Lupo at a conference in San Diego, that neural networks will be more important than the atomic bomb.

So far, the devices on sale hardly justify this statement. "The real situation now is that neural networks are only slightly better than statistical methods," said Smith, a representative of Adaptive Decision Systems. However, taking into account the much greater potential of neural networks, he believes that companies should still continue to work on their practical application. "Investing in these systems is a strategically right step, because the time will come when they will do things that statistical methods cannot." https://mostbets.cz/