Skip to main content

Table 8 Table summarizes the accuracy of ANN, GRU, BILSTM, LSTM, and 1DCNN model for 107 SNPs

From: Eye-color and Type-2 diabetes phenotype prediction from genotype data using deep learning methods

Model, SNPs = 107 Activation Dropout Optimizer Batchsize Epochs Validation Accuracy
ANN Sigmoid 0.3 SGD 1 50 0.3 0.9
Relu 0.2 SGD 1 10 0.3 0.9
Softmax 0.2 Adam 1 10 0.2 0.91
Relu 0.3 SGD 15 50 0.2 0.91
LSTM Relu 0.2 SGD 10 50 0.2 0.9
Relu 0.3 Adam 1 20 0.2 0.9
Relu 0.3 SGD 1 50 0.2 0.9
Sigmoid 0.2 SGD 1 10 0.4 0.91
1DCNN Sigmoid 0.2 Adam 1 20 0.3 0.92
Sigmoid 0.2 Adam 1 50 0.3 0.91
Relu 0.3 Adam 1 20 0.2 0.895
  1. 1DCNN performs well with an accuracy of 0.92%