NumbermlECTS3.0LeveladvancedOverviewMachine Learning (ML) methods allow the analysis of structured data in order to make predictions, classifications, clusterings or recommendations for various purposes. Over the last ten years, Machine Learning has become a key technology for analyzing the growing data volume. Today, the applications of ML methods are omnipresent. Virtually everybody uses them on a daily basis, mostly without noticing them. In business applications they are becoming an important factor for success and therefore a must-have competence for every data scientist.
Machine Learning methods can be categorized into several sub-areas. In this module we discuss a representative selection together with some important general concepts:
Supervised Learning:
- Regression: linear regression and regularization
- Classification: logistic regression, support vector machines, decision trees
- Neural networks
Unsupervised Learning:
- Dimensionality reduction (principal component analysis, etc.)
- Clustering (K-means, Gaussian Mixture Model, etc.)
Model Selection:
- Cross-validation methods
- ML system design (learning curve, information criteria, etc.)
- Anomaly detection
Learning Objectives- The students know basic ML methods, the related algorithms and the most important field of application.
- They are able to implement ML algorithms and apply them to simple but practical problems.
- The students can visualize and interpret the data and the results of the analysis.
- They are also able to quantitatively analyze and evaluate the chosen ML model.
- The students are familiar with a suitable machine learning library and they can use appropriate optimization algorithms.
Previous knowledge- Introduction in Data Science with Python (dsp)
- Advanced Course in Analysis (vana)
- Discrete Stochastic (dist)
- English level B2 (e.g. passed module ten1)
Exam formatContinuous assessment gradeAdditional informationThis module is available as an online course (with additional, graded in-house assignments). 
