Table 1 The main ideas of ANMDA and 6 published methods
From: ANMDA: anti-noise based computational model for predicting potential miRNA-disease associations
Method | Main idea |
|---|---|
ANMDA | Adopts subsampling for noise smoothing and light gradient boosting machine for prediction |
BHCMDA | Uses biased heat conduction-based method to pay attention to specific nodes for prediction |
DBNMDA | Constructs deep-belief network for prediction |
EKRRMDA | Applies ensemble learning and kernel ridge regression on various data subset created by random selection of features for prediction |
FCGCNMDA | Applies fully connected graph convolutional networks for prediction |
MDACNN | Uses auto-encoders for dimensionality reduction and then applies convolutional neural networks for prediction |
WBSMDA | Calculates within-scores and between scores for prediction |