Predicting Popularity of Online Videos Using Support Vector Regression with Gaussian Radial Basis Functions

A wide ranges of recently downloaded videos are economically significant when it comes to the capacity to forecast recent Top-N videos and their aspirations in life. Though numerous efforts have been made to predict video popularity, conventional algorithms do not have the efficacy to foresee top-class viewed videos compared with the whole video set. The truth that almost all videos in the online video system are deceptive is the reason for this occurrence. Thus models should learn to maximize their productivity across the whole video set in an uncomfortable way. The effectiveness of the top-N viewed content, therefore, is essential in most situations in the current study.