Table 4 c-Diadem performance metrics compared with current models
Model | Classification Type | Inputs | Evaluation dataset | Accuracy | AUC | F1 score | Reference |
|---|---|---|---|---|---|---|---|
c-Diadem | Binary (CN, MCI/AD) | SNPs and gene expression data | ADNI test dataset (30%) | 0.6898 | 0.7027 | 0.6898 | This work |
Unconstrained model | 0.5935 | 0.6549 | 0.5935 | This work | |||
c-Diadem | SNPs only | 0.6417 | 0.6702 | 0.6417 | This work | ||
DNN with DEG | Binary (CN, AD) | Blood gene expression | Internal fivefold CV | NA | 0.6568 | NA | [26] |
SNP (deep model) | Binary (CN, MCI/AD) | SNPs | ADNI test set (10%) | 0.66 | NA | 0.53 | [27] |
RPART | Binary (CN, AD) | SNPs | ADNI validation dataset | 0.754 | 0.614 | 0.392 | [28] |