Discriminant analysis and binary logistic regression enable more accurate prediction of autism spectrum disorder than principal component analysis.

Both models were highly significant, as indicated by their respective Chi-square p values (2.66 × 10 –9 and 2.24 × 10 –9 ) and explained more than 45% of data variance as indicated by the corresponding Wilks’ Lambda statistics (Table 1 b).

also in 132,142 other papers