Datasets#

To test the capabilities of XAIoGraphs, it provides a series of datasets via xaiographs.datasets module’s features.

The following datasets are included:

Datset	Rows	Num. Feats	Task
Titanic	1309	8	Binary
Compas	4230	7	Multi-Class (3)
Compas Reality	4230	7	Binary
Body Performace	13393	11	Multi-Class (3)
Education Performance	145	29	Multi-Class (5)

These datasets are accessible in both raw and discretized form, ready for usage by the Explainability and Fairness classes.

The details of these Datasets are shown below:

Note

The original Datasets have been treated to remove outlayers, impute null values, and so on.

Titanic#

The supposedly “unsinkable” RMS Titanic sank on April 15, 1912, during her first voyage after hitting an iceberg. Unfortunately, there were not enough lifeboats to accommodate everyone, and 1502 out of 2224 passengers and staff perished.Individual Titanic passengers’ chances of survival are described in the famous Titanic Dataset.


Source	https://www.kaggle.com/c/titanic
Num Rows:	1309
Num Features	8
Num Targets:	2
function to obtain dataset	`xaiographs.datasets.load_titanic()`
function to obtain discretized dataset	`xaiographs.datasets.load_titanic_discretized()`

Compas#

COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) is a popular commercial algorithm used by judges and parole officers for scoring criminal defendant’s likelihood of reoffending (recidivism). It has been shown that the algorithm is biased in favor of white defendants, and against black inmates, based on a 2 year follow up study (i.e who actually committed crimes or violent crimes after 2 years). The pattern of mistakes, as measured by precision/sensitivity is notable.


Source	https://github.com/propublica/compas-analysis
Num Rows:	4230
Num Features	7
Num Targets:	3 (model) - 2 (reality)
function to obtain dataset	`xaiographs.datasets.load_compas()`
function to obtain discretized dataset (Model)	`xaiographs.datasets.load_compas_discretized()`
function to obtain discretized dataset (Reality)	`xaiographs.datasets.load_compas_reality_discretized()`

Body Performance#

This dataset demonstrates how performance levels change with age and some exercise-related variables.


Source	https://www.kaggle.com/datasets/kukuroo3/body-performance-data
Num Rows:	13393
Num Features	11
Num Targets:	3
function to obtain dataset	`xaiographs.datasets.load_body_performance()`
function to obtain discretized dataset	`xaiographs.datasets.load_body_performance_discretized()`

Education Performance#

The data was collected from the Faculty of Engineering and Faculty of Educational Sciences students in 2019. The purpose is to predict students’ end-of-term performances using ML techniques.


Source	https://www.kaggle.com/datasets/mariazhokhova/higher-education-students-performance-evaluation
Num Rows:	145
Num Features	29
Num Targets:	5
function to obtain dataset	`xaiographs.datasets.load_education_performance()`
function to obtain discretized dataset	`xaiographs.datasets.load_education_performance_discretized()`

< 📚 User Guide