/marriage-prediction

A model that predicts whether a US citizen is married, given multiple factors (occupation, sex, etc)

Primary LanguageJupyter Notebook

Predicting whether a person is married based on a number of facts about them

This model uses an XGBoost decision tree classifier to predict whether a US citizen is married. Practical uses for this include helping one gauge how far behind in one's love life one might be. I do not know how well this generalizes to nationals of other nations, but for US citizens the model achieved around 77% accuracy on its test set, and 75% accuracy on its training set.

I built this to hone my machine learning skills. Later on this may have a frontend attached to it. It may also do other things with the adults.csv dataset. I do intend to use a neural network to predict income on this one.