4  Conclusion

4.1 Summary

The rise of Respiratory syncytial virus hosipitalization rates has been getting us more and more attention. A reliable regression model to predict the trend of the virus is very important.

We have built a model that has a well fit (multiple R square =0.9606 and RMSE=0.16) through polynomial regression analysis in this article. Also, compared the model based on the data with one year span with two year span, we found that using one year-to-date data could help to build a better model for predicting RSV hospitalization rates. Further researches might be conducted to find out if modeling according to one year-to-data is always better to predict some other seasonal diseases, like RSV.

Furthermore, next 3 month (11/14/2022-2/5/2023) RSV hospitalization rates were calculated. The hospitalization rates in the next 3 month could be as high as 9%.

4.2 Limitation

First, our model created from polynomial regression analysis may have a good fit within certain range of the data we selected, but for outside the range of the data, the prediction might not be accurate. To gain a better accurate prediction, people need to repeatedly generate new models.

Second, our data set only includes 58 counties in 12 states from 2018 to now, it might not give us the whole picture of how respiratory syncytial virus evolved and spread in USA. More data is needed to provide a more precise prediction.