for example marks list award with grades. the grades are given to the student if he / she passed all subjects. and then the grading will be given basing on percentage =((total scored marks in all subjects)/(max marks total of all subjects))x100
grade A = 60% to 100%
grade B = 45% to 59%
grade C = 35% to 45%
Hi Veera,
If you have a classification problem where you need to predict if the grade is either A or B or C, then you need to insert each student with his/her grade in each subject (i.e., student is row, subject is column, and the class is either A or B or C). You need to label your data manually as you know students' grades based on the minimum level of success (that you specified) for each of those three categories.
HTH.
Samer
mostly in our area universities and board of academics will give grades on the basis of the percentage of total marks secured out of total max marks of all subjects. but every subject should be get passed(minimum marks of 35% out of max marks of the subject)
If your aim predicting for each student (a) grade and (b) state, then it's a multi-label classification.
Note: Please, next time be more specific and exactly say what you need to predict.
HTH.
Samer
You basically need 3 models that will predict the probability for a student to get one of each grade. The final label would be the one that has the most probability over one student.
You're welcome
- Badges
- Science topic
- Similar topics
- Data Mining
- Classification