School Dropout Prediction 🎓

This tool has been built in order to prevent primary school dropout in high-risk rural regions of Morocco. Student drop out can be identified as a pattern related to socio-economic backgrounds. It is a pressing issue in developing countries and it can be prevented when tackled early. Our analysis focus only on regios whre level of school dropout is high.

Why do we care?

Higher unemployment rate;
increased poverty;
out-of-school students;
increased violence

Rural environment are often neglected by the government, the tool we created allows to raise awareness and easily identify at-risk students.

Link to the app on Streamlit

DB Schema

Key columns

baseline_household

work_activity - Work status of the parents
individual_water_net - Individual water network connection
electrical_net - Electric connection
mobile_phones - If the family have a mobile phone or not
type_housing - The architectural structure of the house

child_math_test_result

digit_recognition_res - Digit recognition test results
number_recognition - Number recognition test results
subtraction_res - Subtraction test results
division_res - Division test resuls
average_math_score - Average score of the 4 math test sections mentioned above

Reference Key

Variable Name	Encoded Numbers	Description
mother_alive	1	Yes
	2	No
father_alive	1	Yes
	2	No
parents_age	-	Age in Years
marital_status	1	Married
	2	Single
	3	Divorced
	4	Widowed
parents_level_ed	1	No education
	2	Religious education
	3	Primary School
	4	Middle School
	5	High School
	6	Higher Education
	7	Professional Training
work_activity	1	Full Time
	2	Part Time
	3	Unemployed
type_housing	1	Adobe/Clay house
	2	Permanent House
	3	Dry Stone
	4	Modern/Concrete house
	5	Other

Dataset