Neural networks and fuzzy logic have merged as useful tools for the calibration of arrays of thin film gas senor. The choice of network parameters is essential for acceptable network performance. Often, choosing network-operating parameters involves a search of many possible candidate networks. When the neural network code is incorporated with other code, such as for data capture, presentation, and control, it is often the case that interdependencies are formed between these code segments. Frequently, enhancements, modifications, and fixes to the code lead to an extensive and time-consuming rewrite of many parts of the software. Thus, the need arise for neural network software modules that can be easily incorporated in application software but whose interface is well defined and whose implementation is entirely separate from the functionality it provides. By providing debugged and proven software modules encapsulating neural network functionality that can be simply inserted into any application, the entire software system can be modularized. These modules can be reused easily, and changing the neural network operating parameters no longer involves a complete software rewrite or even a recompile.