08/10/2020:
9PM: Main work:
* implement relative, k-means grading and compare them and share insights
* tips = gaps finding and gaps dealing.
Points:
* get relative rule and assign grades.
* the problem will be in gaps.
* find gaps heuristic function or DBI formula
* apply to evaluate
* find kmeans algo
* apply kmeans algo for grades
* evaluate clustering quality
* show results to sir.
* do literature review of last research paper.
13/09/2020:
Problem
1. get an excel sheet
2. convert marks to grades
3. apply different techniques
There is a text file with input policy, how many to assign A grades, min limit, max limit, min %, max %
and same for all other grades. A simple setting, a complex settings below.
Done:
1. implemented grade_absolute()
2. implemented graph on excel spread sheet.
2. added git
02/10/2020:
Problem:
1. shayan, muneeb algo:
max marks (93->100%)
make peecentiles
outliers nikal dein (upper, lower)
normal-distr rank
apply k-means
muneeb:
tree of different tech...
7 Oct 2020:
10PM:
Progress:
I did some coding in python. I was trying to implement relative grading. Python skills were weak. So I will revise python in Sololearn app and code tomorrow IA.
Thoughts:
0. Sort on excel sheet. X concept in softdev.
I = No grade
I want to apply these features. Now.
1. grades_relative(marks_list, rule) #marks the list with grades.
1. grades_relative(
marks_list = [[20, 'I'], [35, 'I'], [80, 'I'], [50, 'I'], [30, 'I']],
rule = [[5,'A+'], [5,'A'], [5,'A-'],
[10,'B+'], [10,'B'], [10,'B-'],
[10,'C+'], [10,'C'], [10,'C-'],
[10,'D+'], [5,'D'], [5,'D-'],
[5,'F']])
Later:
1. grades_kmeans(marks_list, rule)
grades_kmeans(marks_list, clusters = 10, grades = [A+, A, A-, B+, B, B-, C+, C, C-, D+, D, D-, F])
Later On:
Load grades, rule from excel sheet or txt file.
#Reference Code:
def sort_marks_column(filename):
data = np.array([[5,2], [4,1], [3,6]])
data = data[data[:,0].argsort()] # sort on column 0
print('data:')
print(data)