deepGraphh

Loading Input

Data Preprocessing

Splitting Data

Model Training

Prediction on Train Data

Prediction on Test Data

Confusion Matrix

ROC Curve

Classification Report

Input Data

Sample File

Query Data

Sample File

Parameters

Mode :

Classification

Regression

This field marks the number of tasks for model

n_Tasks

Value:

n_classes

Value:

The number of classes to predict per task (only used when mode is ‘classification’)

number_atom_features

Value:

The length of the initial atom feature vectors.

n_attention_heads

Value:

Number of attention heads in each GAT layer.

dropout

Value:

The dropout probability within each GAT layer.

predictor_hidden_feats

Value:

The size for hidden representations in the output MLP predictor.

predictor_dropout

Value:

The dropout probability for the output MLP predictor.

alpha

Value:

A hyperparameter in LeakyReLU, which is the slope for negative values.

Self loop

ELU Activation

Layers in Graph Attentive Network:

Aggregation mode:

Flatten

Mean

Note: Please perform k-fold Cross-validation only once the final Model Parameters are finalized. Please note live tracking is disabled for this process.

Cross-validation:

5-fold Cross-validation

10-fold Cross-validation

None