The invention is a method of allocating a computer to service a request for a data set in a system having a plurality of computers. The method is implemented on a neural having at an input layer having input nodes and an output layer having output nodes, where each output node is associated with a specific computer. Connecting the input nodes to the output nodes are weights w(j,k). Each output node is associated with a computer in the system, and the inputs to the input nodes are dependent upon the number of requests for specific pages.