Abstract:
YouTube is a global online video sharing and social media platform currently owned
by Google. Trending videos represent the content that is gaining viewership over-a
certain period. While popular videos get a huge number oflikes and views with time.
Trending video analysis has not been properly analysed yet despite their importance.
The study is conducted to analyse features to determine the importance of variables
for the trendiness of a video and focuses on how these features help a video to trend
on YouTube. The study also includes learning different models on the dataset but will
also conduct a comparative analysis between yearly datasets. Our work also includes
classification models known as Hard Voting, Gradient Boosting, Gaussian Naive.-
Bayes, Logistic Regression, and Random Forest to determine which model suits better
for prediction
