Helsinki University of Technology → Department of Computer Science and Engineering →
Laboratory of Computer and Information Science → Teaching →
T-61.3050 Machine Learning: Basic Principles → 2007 → Grading principles

Grading principles 2007 - T-61.3050

To pass the course both the examination and the term project must be passed.

The 2007 examination, if passed, is graded on a scale from 0 to 1 (grade may be fractional). The smallest passed grade is 0. The grade is a monotonic, but not necessarily linear, function of total points from the examination. We have used E=(PT-15)/15, where PT are the total points from the examination (smallest passed total points being PT=15, highest PT=30) - the exact formula may however change depending, for example, on the difficulty of the examination. Denote the examination grade by E.

The 2007 term project, if passed, is graded on a scale from 0 to 1 (grade may be fractional). The smallest passed grade is 0. Denote the term project grade by T.

The 2007 problem sessions are graded on a scale from 0 to 1 (grade may be fractional). The grade of 0 corresponds to no points from the problem sessions. Denote the problem session grade by P.

The course grade is then given by min(5,floor(1+4*E+2*T+P)).

Guidelines for grading of the answers

Some general guidelines for grading of the examination and term project answers are given below (which are quite similar to the grading criteria of the Finnish high school final examination, "reaalikoe"). Notice that in addition to having a correct factual content a good answer should be, among other things, understandable for the examiner.

Signs of a strong answer:

• The answer is structured and the factual content is correct and relevant.
• There is sufficient amount of essential information; the length of an answer or the number of details are not merits in themselves.
• Causes and consequences are discussed appropriately from different viewpoints.
• All claims made are substantiated.
• The answer indicates readiness to independently process and apply the related knowledge and skills.
• Any given source material is used appropriately.
• The student relates the knowledge presented in an answer to the larger context.
• A clear distinction is made between facts, substantiated claims and opinions.
• The editing is to the point, clear, logical and exact; appropriate notations and conventions (in mathematical derivations, pseudocode etc.) are observed.

Signs of a weak answer:

• The answer contains factual errors.
• The ideas are presented un-clearly or inaccurately.
• The presented knowledge indicates that the student has misunderstood the problem, or the presented facts are otherwise irrelevant.
• The answer is based only on opinions.