Description
| Abstract: | This paper determines the statistical properties of correspondence analysis (CA) when compared to classical methodologies under constraints imposed by experimentation, but not by the modelling and probability assumptions generally imposed. Specifically, CA will be compared to classical techniques when applied to speech pattern data which are being used to discriminate between normal elderly subjects and those suffering from senile dementia of Alzheimer's disease. |
| Item Description: | Offprint: American Statistical Association. Proceedings of the Statistical Computing Section. |
| Physical Description: | 12 leaves, 3 unnumbered leaves : illustrations ; 28 cm |
| Bibliography: | Includes bibliographical references (leaves 11-12). |