Traffic Volume Forecasting Tool Simulates Human Memory

Emulating Human Memory
The Spinning Network (SPN) approach is based on the understanding that human memory is an essential part of human intelligence. Thus, it seeks to enhance artificial intelligence by emulating several features of human memory. These include the imprecise nature of information received, the association of ideas, and the improvement of information retrieval through an investment of time and effort.

Organizing the Database
SPN continually observes data and organizes it into memory. Conceptually, this approach consists of concentric data storage “rings” that spin asynchronously, with the outmost ring spinning fastest and having the greatest storage capacity. Each ring has an input window for receiving new data, and a “To Next Ring” window which merges similar data, deletes it from the current ring, and moves it to an inner ring.

In output mode, new inputs are compared to existing data. Each ring supplies its own output, and the most similar data is selected and output by the SPN.

Fast & Efficient
SPN offers several advantages over existing approaches to traffic flow forecasting. Unlike the nearest neighbor approach, for example, SPN uses a relatively small database that significantly improves runtime speed. Also, in contrast to the neural network approach, SPN does not require a time-consuming training period.

File Number: 349 

Web site: http://www.uvminnovations.com/it.html

Other Information:

Commercialization
Additional applications for SPN data modeling include weather forecasting, market forecasting, human memory simulation, and online data clustering. The licensed product can be delivered as a customized computer program, software module, or web-based application.