Go viral or not?

I used XGBoost to analyze the Online News Popularity Data and predict the popularity of online news articles. The aim is to construct a model and extract feature importance. The research question is as follows: Can gradient boosting be constructed based solely on the research data?

H0: Gradient boosting cannot be made from the Online News Popularity dataset.
H1: Gradient boosting can be made from the Online News Popularity dataset.