HSC exam mark predictor/predictive model based on school data
- Simon A
- May 29
- 1 min read
An initial attempt at using historical data as a predictor of future HSC exam results from the school assessment marks. This uses data from different subjects over various years and employs a simple linear regression and also Random Forest (Machine Learning) models to predict a future HSC result based on School Assessment results.
This is school-specific and is a general prediction for results. This could be used to help in try to set a goal as a student and for teachers with additional data to try to create more accurate moderation. The model in various subjects is either strong, moderate or weak shown by RMSE, r and p- value. It also shows the linear regression with marks, and where you mark would be placed. Addionally, there is a residual plot, closer together along the 0 would be a good fit. if your mark falls in a high residual difference it may have some noise.
Marks on the lower end do tend to jump in some subjects and this could create some inaccuracies. Note that this is for informational purposes and the models are at times mostly moderate, with the exception of stronger models in English and Physics and a few others.
