Exam
Format and grading
The course ends with an exam. The exam will be on campus and will contain both pen and paper questions as well as programming questions. The exam will be done on a computer using Inspera Assessment system. See further information here Links to an external site. regarding how it works to do a digital exam on campus.
The exam will consist of a combination of automatically corrected multiple choice questions and free text questions and problem solving assignments. Both pen and paper questions as well as programming quesations will be included. The solutions to the exam questions will in most cases be entered digitally directly in Inspera. For questions where the expected solution contains mathematical expressions, these solution will be handed in on paper as in a normal written exam.Format
1RT003: 5 problems, total 50 points
1RT705: 4 problems, total 40 points
Preliminary grade limits
1RT003
Grade 3: 20hp
Grade 4: 30hp
Grade 5: 40hp
1RT705
Grade 3: 16hp
Grade 4: 24hp
Grade 5: 32hp
Exam time
1RT003: 5 hours
1RT705: 4 hours
Aiding material: One (1) hand-written sheet of paper with notes and formulas (A4, front and back).
Provided material: (provided, listed as digital resources in the e-exam):
- Numpy documentation Links to an external site.
- Scipy documentation Links to an external site.
- Matplotlib documentation Links to an external site.
- Lecture slides Download Lecture slides
- Formula sheet for the Gaussian distribution Download Formula sheet for the Gaussian distribution
- Christopher M. Bishop, Pattern Recognition and Machine Learning
, Springer, 2006.
- Lindholm et al, A First Course for Engineers and Scientists Links to an external site. , 2021 Machine Learning -
- Programming tool Links to an external site.
Course grade
To pass the course, you need to pass the mini-project, the peer-review assignment, the lab (only for 1RT003) and the exam. The course grade will then be the grade you got on the exam.
Demo exam
This is the first year the course ends with a written exam (previous years it was an oral exam). We have therefore made a demo exam available so you can familiarize yourself with the format.
It is strongly recommended that you look at the demo exam before the real exam to familiarize yourself with the graphical interface for doing the programming assignments and how to copy code and the requested plots into the exam. If you have any questions about the procedure and format, please identify them before the exam. You can ask them here. The proportion of assignments related to the different parts of the course might differ as well as the proportion of question types, such as programming, pen and paper, free-text and automatically corrected assignments.
Old exams
- 2022-08-24 Exam 1RT003 Links to an external site. (pdf Download pdf), Exam 1RT705 Links to an external site. (pdf Download pdf), Solutions Download Solutions
- 2021-12-20 Exam 1RT003 Links to an external site. (pdf Download pdf), Exam 1RT705 Links to an external site. (pdf Download pdf), Solutions Download Solutions
- 2021-10-22: Exam 1RT003 Links to an external site. (pdf Download pdf) , Exam 1RT705 Links to an external site. (pdf Download pdf), Solutions Download Solutions
- Demo exam: Exam 1RT003 Links to an external site., Exam 1RT705 Links to an external site., Solutions Download Solutions
Re-exam
Re-exams will be offered during the re-exam periods in December and August.
Registration
You register to the exam via ladok no later than 12 days before the exam.
PhD-students
PhD students enrolled in FNT0092 and FNT0204 will take the same exam as 1RT705 and 1RT003, respectively. FNT0092 and FNT0204 are only graded with U/G. To get grade G on the course, grade 4 on the exam is required for PhD students.
PhD students register to the exam by sending an email to it-kansli@it.uu.se including personal number and the course code of the exam (1RT705 if you are resisted on the 5hp version or 1RT003 if you are resisted on the 7.5hp version). This should also be done no later than 12 days before the exam.