Biomedical Data Science
(formerly: Algorithmic Bioinformatics)
Teaching

Lecture: Data Mining and Machine Learning in Bioinformatics (WS 2018/19)

Location: B-IT, room 0.107

Time: Mo, 9:30 - 11:00

Exercises: NONE 

Prerequisites: Bachelor Computer Science or equivalent qualification

Content: This lecture gives a broad overview about data science methods, which are frequently used in life science informatics. The goal is to understand the explained methods and being able to apply them correctly in the life science context. More detailed, the  the following topics are covered:

  • Statistical Basics of Data Mining
    • Descriptive statistics and statistical plots
    • Probability distributions
    • Statistical hypothesis testing
    • Linear models, logistic regression
    • Principal Component Analysis
  • Cluster Analysis
    • Basic methods: hierarchical clustering, k-means
    • Gaussian Mixture Models, consensus clustering, Non-negative matrix factorization
  • Classical Supervised Machine Learning Methods
    • Principles of supervised learning: bias-variance trade-off, regularized risk minimization
    • Penalized generalized linear models: lasso, ridge, elastic net
    • Random Forests
    • optionally: Gradient Boosting
  • Deep Learning Approaches
    • (Convolutional) Deep Neural Networks, Recurrent Neural Networks
    • (Variational) Autoencoders, Generative Adverserial Networks
  • Optionally: Bayesian Networks

Prerequisites for getting credits are:

  • passing of exam

Seminar: Machine Learning Methods in the Life Sciences (SS 2019)

Location: 3.110, b-it

Time: Mo, 9:30 - 11:00am

Prerequisites:

  • B.sc. Computer Science or equivalent qualification
  • Knowledge about basic supervised and unsupervised machine learning techniques, e.g. neural networks, Random Forests, SVMs, PCA

Credits:

  • presentation of one chosen topic
  • regular attendance

Topics, presenters and dates