Machine Learning Engineer Nanodegree

Capstone Project

Project: Classifying Households

This project uses supervised learning techniques for classification of households into two groups for payments via cash transfers

Install

This project requires *Python 3.6. ** and the following Python libraries installed:

• NumPy
• Pandas
• matplotlib
• scikit-learn
• XGBoost
• Jupyter notebook / iPython Notebook

Data

The dataset consists of 100,000, with 18 features. The dataset can be downloaded from this link on github

Features

• Gender: Male, Female
• Age: continuous.
• Attended_school: Currently attending, Never, Yes but not now, SKIP
• Work_last_7days (had work at time of registration ?): SKIP, Worked for pay, On leave, Sick leave, Worked on own/family, business, Herding-Worked on own/family agri. holding, Seeking work, Doing nothing, Retired, Homemaker, Full-time, student, In capacitated, Other
• Chronic_illness: 0 - false, 1- true, 2-skip
• Disabled: 0- false, 1- true
• YearsChronicallyIll: continuous
• WealthGroup: Better Off, Middle, Poor, Very Poor
• Resident_Provider: SKIP, All live here all of the time, Some live here while others migrate, All migrate together-fully mobile, Inside this household, Outside household but inside compound-settlement, Outside this compound-settlement
• Polygamous: discrete
• Children_Under_15: discrete
• kids_under_15_in_settlement: discrete
• wives_in_settlement: discrete
• spouses_outside_hh: discrete
• Toilet: Flush toilet, VIP latrine, Uncovered pit latrine, Covered pit latrine, Bucket/pan, Bush-None, Other
• Drinking_water: Piped water inside dwelling, Piped water into plot/yard, Public tap, Tube well/borehole with pump, Protected dug well, Protected spring, Rainwater collection, Unprotected dug well/springs, River, Lake, ponds or similar, Water truck/vendor, Bottled water, Other
• Land_Ha: discrete

Target Variable

• HHGroup: categorical (True, False)