59 lines
2.5 KiB
JSON
59 lines
2.5 KiB
JSON
{
|
|
"_comment": "Model configuration template for the classification model builder plugin.",
|
|
"model_name": "ExampleClassifier",
|
|
"_comment": "A descriptive name for your model.",
|
|
"model_type": "RandomForestClassifier",
|
|
"_comment": "The type of classification model to use (e.g., RandomForestClassifier, LogisticRegression, SVM).",
|
|
"data_path": "data/training_data.csv",
|
|
"_comment": "Path to the CSV file containing the training data.",
|
|
"target_column": "target",
|
|
"_comment": "The name of the column containing the target variable.",
|
|
"features": [
|
|
"feature1",
|
|
"feature2",
|
|
"feature3",
|
|
"feature4"
|
|
],
|
|
"_comment": "List of column names to use as features. If empty, all columns except the target_column will be used.",
|
|
"hyperparameters": {
|
|
"_comment": "Hyperparameters specific to the chosen model type.",
|
|
"n_estimators": 100,
|
|
"_comment": "Number of trees in the random forest (example for RandomForestClassifier).",
|
|
"max_depth": 10,
|
|
"_comment": "Maximum depth of the trees (example for RandomForestClassifier).",
|
|
"learning_rate": 0.1
|
|
"_comment": "Learning rate for gradient boosting models (example for GradientBoostingClassifier)."
|
|
},
|
|
"training_parameters": {
|
|
"_comment": "Parameters related to the training process.",
|
|
"test_size": 0.2,
|
|
"_comment": "The proportion of the data to use for testing.",
|
|
"random_state": 42,
|
|
"_comment": "A random seed for reproducibility.",
|
|
"stratify": true
|
|
"_comment": "Whether to stratify the target variable during train/test split."
|
|
},
|
|
"evaluation_metrics": [
|
|
"accuracy",
|
|
"precision",
|
|
"recall",
|
|
"f1-score",
|
|
"roc_auc"
|
|
],
|
|
"_comment": "List of evaluation metrics to compute on the test set.",
|
|
"model_save_path": "models/example_classifier.pkl",
|
|
"_comment": "Path to save the trained model.",
|
|
"feature_importance": true,
|
|
"_comment": "Boolean value to toggle feature importance calculation.",
|
|
"preprocessing": {
|
|
"_comment": "Configuration for data preprocessing steps.",
|
|
"handle_missing_values": "impute",
|
|
"_comment": "How to handle missing values (e.g., 'impute', 'remove', 'none').",
|
|
"missing_value_strategy": "mean",
|
|
"_comment": "Strategy for imputation (e.g., 'mean', 'median', 'most_frequent').",
|
|
"scale_features": true,
|
|
"_comment": "Whether to scale numerical features using StandardScaler or similar.",
|
|
"feature_scaling_method": "standard"
|
|
"_comment": "Method to use for feature scaling ('standard' or 'minmax')."
|
|
}
|
|
} |