Research Of Algorithm For Ratings Forecasting And Software Development

With the development of modern management, the funds, which is appropriated by government to the TV media is much less than before. So it is very important for the TV media to get more incomes with its limited resource. And it is necessary to be researched. This thesis studies how the TV media can predict its ratings reasonably and effectively.The article firstly introduces the current research situation of the ratings forecasting, significance for research and raises the issue for forecasting ratings. We then refer to the principles of data mining to predict the ratings , including the algorithm of decision tree theory and beyes classification theory ,constructing the model to deal with the actual problem. For classification of data mining can only classify the ratings, we further solve the problem in accordance with the characteristics of the time series of ratings based on ARMA model. The simulation results and analysis show that the models in the process of TV media ratings prediction is feasible and effective. Through a lot of research for the theories of Data Mining classification and time-series to predict the ratings, We design the software system on forecasting for the TV ratings. According to the principles of software engineering , the system is robust, reliable, scalable and efficient.