The viewer is disabled because this dataset repo requires arbitrary Python code execution. Please consider removing the loading script and relying on automated data support (you can use convert_to_parquet from the datasets library). If this is not possible, please open a discussion for direct help.

YAML Metadata Error: "configs[0]" must be of type object

NBFI

The NBFI dataset from the Kaggle. Client default prediction.

Configuration Task Description
default Binary classification Has the client defaulted?

Usage

from datasets import load_dataset

dataset = load_dataset("mstz/nbfi")["train"]

Features

Feature Type
income float32
owns_a_car bool
owns_a_bike bool
has_an_active_loan bool
owns_a_house bool
nr_children int8
credit float32
loan_annuity float32
accompanied_by string
income_type string
education_level float32
marital_status float32
is_male bool
type_of_contract string
type_of_housing string
residence_density float32
age_in_days int32
consecutive_days_of_employment int16
nr_days_since_last_registration_change int32
nr_days_since_last_document_change int32
owned_a_house_for_nr_days int32
has_provided_a_mobile_number bool
has_provided_a_home_number bool
was_reachable_at_work bool
job string
nr_family_members int8
city_rating int8
weekday_of_application int8
hour_of_application float32
same_residence_and_home bool
same_work_and_home bool
score_1 float32
score_2 float32
score_3 float32
nr_defaults_in_social_circle int8
inquiries_in_last_year float32
Downloads last month
0
Edit dataset card