The Statistical Analysis based on YouTube Channel Dataset

Using machine learning analysis methods could provide an important reference basis for the company's precision marketing. This paper used a series of machine learning analysis methods on the YouTube channel dataset. These algorithms include linear regression, correlation analysis, k-means cluster analysis, and social network analysis. Based on the results of these algorithms, it can be concluded that a moderate negative association exists between length-view, length-ratings, and length-comments where length is the time length for a specific video while views-ratings, views-comments, ratings-comments are positively associated.