We hear about celebrity breakups almost every day. I’m curious about if celebrities have a higher divorce rate than the general population? If so, how high their divorce rate could be? How long are their marriages last? What’s the key reason to drive their divorces? Is it predictable?
This post will dive into the Hollywood celebrities data parsed from Wikipedia and try to provide insights on the celebrities marriage and divorce by analyzing these data.
Three questions will be answers in this post:
1. What is the divorce rate amongst U.S celebrities in Hollywood compared to the general U.S population?
2. How long is celebrity marriage last?
3. Which feature has the most impact on celebrity divorce?
Question 1: What is the divorce rate amongst U.S celebrities in Hollywood compared to the general U.S population?
According to 2017 US Census data, about 23% of married couples in the United States divorce. Among them, the divorce rate for male is 20%, and for female is 26%.
However, the average divorce rate of Hollywood celebrities is 52%, which is 2 times higher than the general population. Female has 62% divorce rate, while the divorce rate of male is 50%.
Question 2: How long is celebrity marriage last?
Hollywood is known for its quickie marriages and divorces. According to Wikipedia, the length of their marriages can be as short as hours to days. In this data analysis, we found the median value for celebrities’ first marriage is 6 year. There is no significant difference between male and female.
Interestingly, when considering the remarriage, female takes more time than male. Female takes 7 years for their remarriage, while male only takes 3 years. This suggests that women are less likely to be remarriage or consider more than men do.
Question 3: Which feature has the most impact on celebrity divorce?
To answer this question, we used the random forest algorithm to calculate the feature importance from the data. The result shows the most important features to lead to their divorce are their current age and the age at their first marriage. We can clearly see that there is a second distribution between age 80 and 160 in the divorce population. According to a recent report, the older generation had unusually higher rates of divorce than the Millennials or Generation X. So it is not surprising that age can affects celebrity divorce.
Another interesting finding is the age of the first marriage can significantly affect celebrity divorce. people who married at a younger age is most likely to divorce.
In this project, we applied the data to two machine learning models (Logistic Regression and Random Forest) with the goal to predict celebrity divorce. Although the data was wrangled from Wikipedia, we can still get about 70% accuracy from both models. It is possible to predict celebrity divorce. Applying more features and using more clean data will help optimize the model and increase the accuracy of prediction.
You can have a closer look at the data and code in my GitHub repository if you are interested.