From block 1704 control proceeds to block 1706 in which a margin of error is calculated based on comparing the generated output value to an expected output value, wherein the expected output value is generated from an indication of a predetermined function based at least on a number of I/O operations that are waiting for a resource and a number of available resources. Control proceeds to block 1708 in which an adjustment is made of weights of links that interconnect nodes of the plurality of layers via back propagation to reduce the margin of error, to improve a determination of the number of resources to allocate to the interface.

It should be noted that the margin of error for the machine learning module may be computed differently in different embodiments. In certain embodiments, the margin of error for training the machine learning module may be based on comparing the generated output value of the machine learning to an expected output value. Other embodiments may calculate the margin of error via different mechanisms. A plurality of margin of errors may be aggregated into a single margin of error and the single margin of error may be used to adjust weights and biases, or the machine learning module may adjust weights and biases based on a plurality of margin of errors.

Therefore, FIGS. 1-17 illustrate certain embodiments, in which a machine learning module 106 is used to determine balance a plurality of resources among a plurality of interfaces of a storage controller 102. Training mechanisms for the machine learning module are also discussed in FIGS. 1-17.