This project was made for the DTU course: "Data Mining using Python" (number 02819). Click on the following link to see the details using this system: this link. If you think this project is cool and want to improve upon it, shoot me an email on firstname.lastname@example.org. A little info on the different pages of the project:
- Home: Contains easy access to the courses page, by using either department, course type or course responsible to see courses. This is smart if you've found a lecturer you really like - and whats to take more of his courses, or if you're just missing 5 points in your technological specialization courses, you can quickly find them and then sort by for example evaluations.
- Courses: If you just access the page directly (meaning no querystring from the home page), a list of all available courses will be loaded, from which you can easily sort by the parameters most important to you. One very important feature of this page is that you can double click a row (course) and see further informations about it in a modal window.
- Grades: It should be noted, that the grade in the course overview and the toplists is a Bayesian average given by, calculating a so-called credible grade:
Gcred = (G*s + C*m) / (s+m)
Here G is the average grade in the course, s the number of students, C the average grade from all grades registered on the course base and m some minimum of students (here 15). This avoids courses with an average of grade of 12.0 while only having three students enrolled, being shown as 12.0.
- Toplists: Toplists (and bottomlists) by grade, evaluations and pass rate.
- Analysis: A proper name for this would have been visualizations. Many different patterns could be extrapolate from the data:
If you're interested in the data for analysis, you're welcome to contact me.
- The number of students, exams being attended, courses are growing for each year.
- DTU seems to be focusing more on masters students then in the past.
- The grade on masters studies is significantly higher than it is on bachelor studies - maybe a sign of student maturity, maybe because of more project work on the masters.
- The biggest department is DTU Compute, even though it doesn't have the most courses.
- 00 is given more than 02. 12, 10 and 7 are given more than 00, 02 and 4. The average (from all the grades) is 6.8, so highly likely 00 is pulling it down.
- The grade has risen over time. This could be due to the 7-point scale being introduced, making 10's and 12's more frequent.
- Exams participants for course levels are almost equal between BSc, MSc and diplom, but when looking at the number of MSc courses offered, it's 2.5 times that of BSc courses offered.
- Reporting sick for an exam happens very seldom. Around 1‰ of participants do this.
- Most courses are below 20 students, but some very large courses are pulling this average up to around 35 students per class.