def _numeric_column_normalized(column_name, normalizer_fn):
    return tf.feature_column.numeric_column(column_name, normalizer_fn=normalizer_fn)

# Define your feature columns
def create_feature_cols(features, use_normalization):
    """Create feature columns using tf.feature_column.

    This function will get executed during training, evaluation, and serving."""
    def normalize_column(col):  # Use mean, std defined below.
        return (col - mean)/std
    normalized_feature_columns = []
    for column_name in features:
        normalizer_fn = None
        if use_normalization:
            column_params = normalization_parameters[column_name]
            mean = column_params['mean']
            std = column_params['std']
            normalizer_fn = normalize_column
    normalized_feature_columns.append(_numeric_column_normalized(column_name,
                                                              normalizer_fn))
    return normalized_feature_columns

NUMERIC_FEATURES = ['housing_median_age', 'total_rooms', 'total_bedrooms',
                    'population', 'households', 'median_income']
feature_columns = create_feature_cols(NUMERIC_FEATURES, use_normalization=True)