The prediction of Digital Political Commercial Evaluation Using Improved Linear Regression

Analysis of digital advertisements for politicians with an age factor using linear regression algorithms versus decision trees to determine improved accuracy. The study gathered 116 samples from various internet sources using a 0.05 percent threshold, a confidence level of 95 percent interval, a mean and standard deviation, and an 80 percent G-power. Because the Linear Regression algorithm has already achieved an accuracy of 95.25 percent in forecasting political advertisement analysis, the DT approach must achieve greater accuracy in this research. In this study, the DT algorithm was used to analyzed political ads, yielding a 95.00 percent accuracy with a significant value of 0.007 (p0.05) for two-tailed tests with a confidence level of 95 intervals. The DT approach outperforms the Linear Regression algorithm within the constraints of this experiment.