The rapid prediction of playing index in accumulation on video websites

In order to avoid the problems in traditional forecasting methods which demand too much of various data types and have difficulty in training models, this paper proposes two rapid prediction methods which are called “One by One Comparison” and “Regression as a Whole”. By using the two methods, as long as you get playing index in the first few days of a TV drama, the total playing index accumulation in certain days in future will be predicted fast. By the experiment, the predicting precision of method ”Regression as a Whole” has significant improvement than “Time Series Analysis”, “Curve Fitting” and “One by One Comparison”, so it can be a fast and easy way to predict the playing index accumulation of one TV series.