News
Updates from the Exam Inspection in CMS
Written on 18.08.2026 16:58 by Tim Bauerschmidt
Dear students,
We have now processed all claims from the exam inspection and updated the points in CMS. The results should now be reflected on your personal status page.
We would like to clarify the grading of the exercise on deriving kernelized logistic regression. The exercise required starting from the linear model in feature space, f(x) = wᵀφ(x), and showing how the kernelized formulation follows from it. Note that the Representer Theorem discussed in Lecture 15, slide 11, works in the RKHS and therefore assumes the kernelization that the exercise asked you to derive (see, e.g., Lecture 15, slide 6). Consequently, solutions that did not explicitly use the feature map φ did not receive points for this part.
Best,
ML Team