Skip to content

AntiAntiexe/grade-predictor

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

27 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

The Nostradamus

The Nostradamus is an "accurate" grade predictor. Based of a data set of over 1,000 grades. The Nostradamus can calculate a person's third grade based on their 1st and 2nd. This was all made for a school assignment.

It uses CustomTkinter to create a modern Ui. I opted for the use of CTk instead of Tkinter, even though Tk has more infrastructure built already. But CTk is slowly developing and I think it was the right choice to make.

When someone first reads this project then they just assume that I average the two results. However, this app actually uses a linear regression model from scipy stats. This module imports specific linear regression statistics. These statistics can create the default linear equation, y = mx + c. Using this equation, you can predict certain values, such as input the x value to calculate the y value and vice versa.

Finally, I used matplotlib to display the graph on the application. Displaying the graph will help the user to visualize the data.

For further improvements I will probably change the way that I have normalized the data. Thorugh my research in neural networks, I have found that the normalization formula that I chose:

image

Is not accurate enough for the purpose of predicting grades so to imporve this I have decided to use the softmax function.

image

This function is far more accurate at normalizing data between 0 and 1. This is because of the use of exponents. And as of the latest Numpy version. Calculating with exponents is quite simple.

Getting Started

Follow these steps to start using this app:

  • You can simply download the folder.
  • Make sure you have the modules needed:

    customTkinter, CTk Message Box, Scipy, Matplotlib and Math

  • Finally, run the nostradamus.py file and enjoy!

About

a grade predictor that can predict grades

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages