model overfitting

[tkisss]

Hi, I noticed that CellOracle utilizes regression models like BaggingRegressor to fit gene expression based on transcription factor expression without dividing the data into training and testing sets. Can I infer that in this scenario, the model is prone to overfitting? Does this strategy provide any benefits for gene expression regulation?

if scaling:
    data = gem_scaled[reg_all]
else:
    data = gem[reg_all]
label = gem[target_gene]

try: # For sklearn version 1.2 or later
    model = BaggingRegressor(estimator=Ridge(alpha=alpha,
                                                 solver=solver,
                                                 random_state=123),
                                n_estimators=bagging_number,
                                bootstrap=True,
                                max_features=0.8,
                                n_jobs=n_jobs,
                                verbose=False,
                                random_state=123)
 except: # For old version of sklearn
    model = BaggingRegressor(base_estimator=Ridge(alpha=alpha,
                                                  solver=solver,
                                                  random_state=123),
                                n_estimators=bagging_number,
                                bootstrap=True,
                                max_features=0.8,
                                n_jobs=n_jobs,
                                verbose=False,
                                random_state=123)
model.fit(data, label)