We are interested in the divorce dataset because of its societal importance.

We learned how to apply machine learning models, like Artificial Neural Networks and Logistic Regressions to make predictions from data. We then applied that to predicting marriages and divorce based on survey results.

We faced challenges regarding the large number of questions and small number of samples, but we overcame this using qualitative grouping techniques to create a personality summary based on the survey answers.

We then found that these personality traits are highly predictive of divorce, allowing %13 of the amount of input data to have the same predictive capability.

Built With

Share this project: