CGPA prediction model

[armaf002]

Uploaded a complete CGPA prediction model notebook

[Sammybams]

Thank you very much @armaf002 for your contribution. However, we will not be able to merge your pull requests as it doesn't follow the guidelines already stated in the starter notebook. There are different grading systems for different schools, so it was wrong to remove the data point that had a grading system of 10.

Also, the prediction is based on the normalised CGPA. So the CGPA is divided by the grading system to make it a minimum of 0 and a max of 1. With that, the grading system by the person cannot be an affecting factor. It shouldn't also be a metric to train the model as is a measuring metric, not a feature. This also means that your visualization of the distribution of the CGPA was wrong because grades of students that have a grading system of 4 operated will have their CGPA capped at 4 and hence the visualization tells a false story for CGPAs above 4.0.

So in all visualizations and predictions for this project, the normalised CGPA is what is being worked with. Thank you.

[armaf002]

Oh alright, thank you