| Course log sf2935
Supporting material and lecturers' slides will be given on this page.
- Guest Lecture Erik Aurell: Model learning using (many)
- Protein structures
- Supplementary material: exponential families
- Additional material:
Christoph Feinauer, Marcin J. Skwark, Andrea Pagnani, Erik Aurell, Improving contact prediction along three dimensions PLoS Comput Biol 10(10):
- Question 5 to be answered in submitted Summary: Explain how the exponential family of probability distributions is related to this problem.
- Question 4 to be answered in submitted Summary: Explain the mean field method of approximation for contact prediction
- Lecture 18: Random Forests
- Lecture 17: Unsupervised learning part 2
- Lecture 16: Unsupervised learning part 1
- Lecture 15: an exercise class on pp. 81-87 in the slides of
- Lecture 13-14: support vector machines;
- Sections 3-5 of the linked tutorial as extra material
- Lecture 11-12
- Lecture 9-10: Bayesian Learning
- Paper: Approximate Bayesian Computing
- Question 3 to be answered in a submitted Summary: Explain the basic ABC = Approximate Bayesian Computation for Bayesian model evidence as found in the linked paper:
- Lecture 8.: Guest lecture by Lukas Käll. In addition there are two connected papers I-II below
- Lecture 7.pdf.
- Lecture 6.pdf.
- Lecture 4.pdf.
- Lecture 3.pdf.
- LINEAR PROGRAMMING DISCRIMINANT, (LPD) : pdf.
- Lecture 2.pdf.
- Lecture 1. pdf .
- BIAS-VARIANCE TRADEOFF : link .
- section 6.5.3. in A.Blum, J. Hopcroft and R. Kannan: Foundations of Data Science, 2016 (pdf) : click .