def train_model(dataset: pd.DataFrame, data_class: str) -> JoblibSerializedFile:
x_train, _, y_train, _ = train_test_split(
dataset[dataset.columns[~dataset.columns.isin([data_class])]],
dataset[data_class],
test_size=0.33,
random_state=42,
)
model = GaussianNB()
model.fit(x_train, y_train)
model.feature_names = list(x_train.columns.values)
fname = "/tmp/model.joblib.dat"
joblib.dump(model, fname)
return fname
