A vehicle occupant classification system categorizes vehicle occupants into various classes such as adult, child, infant, etc. to provide variable control for a vehicle restraint system such as an airbag. The classification system utilizes sensors that are installed in various locations in the vehicle. The sensors are used to generate a three-dimensional profile for the vehicle occupant. Various factors can affect the accuracy of this three-dimensional profile. Fuzzy logic is used to reduce some of the inaccuracies by providing multiple decision levels for various stages of the classification. Inaccuracies are also caused by sensors shifting within the system from their original position. This condition creates offset and the system evaluates this offset and generates a correction factor to provide a more accurate three-dimensional profile. Electrically erasable programmable read-only memory is used to reduce complications and inaccuracies associated with seat occupant weight sensors that have mounting configurations that vary depending upon the vehicle.