The results of the classification were compared with the actual surface morphology based on high-resolution orthophotos, aerial photos, and field surveys.
The relative importance of the input features was also compared and the features that contributed the least to the classification, were eliminated to decrease the complexity of the networks.
At the 1992 Paralympics, included on the programme for the classification were the Super G and Downhill disciplines.
Discriminant function and multinomial regression analyses allowed us to convert this mathematical construction into a clinically useful tool in which the two major contributors to the classification were bulbar or limb onset and diagnostic delay.
For example, if the classification were operating at chance level, one would expect a mean normalized rank accuracy of 0.50, indicating that the correct word appeared on average between the 30th and 31st position in the classifier's output of a ranked list of 60 items.
Only 2 modifications to the classification were required.
