|
Lehrende: Thomas Heil
Veranstaltungsart:
Seminar
Orga-Einheit: Corporate Management & Economics
Anzeige im Stundenplan:
Quan. M II ML
Semesterwochenstunden:
3
Credits:
6,0
Hinweis: In Ihrer Prüfungsordnung können abweichende Credits festgelegt sein.
Standort:
Campus der Zeppelin Universität
Unterrichtssprache:
Englisch
Min. | Max. Teilnehmerzahl:
8 | 22
Prioritätsschema: Standard-Priorisierung
Inhalte:
+++++ Please be aware that in Fall Semester 2021 this course will be offered for the last time +++++
This course aims to provide students with basic and advanced knowledge in quantitative methods. Quantitative Methods II primarily focuses on the rapidly emerging field of machine learning. Whereas classical quantitative methods specialize in recreating reality by employing predefined models, machine learning offers the ability to let machines (computers) learn from data. The advocated methods should give the students the ability to apply the course content to their own research projects.
The first part of the course consists of lectures and practical applications using Python (no previous knowledge of Python is required. The course additionally aims to provide introductory knowledge in the programming language Python.)
Part 1 : Introduction: The Problem of Learning
- What is Artificial Intelligence?
- What is Machine Learning?
- What is Statistical Learning?
- What is learning from data?
- Training vs. Test sets
- Supervised vs. Unsupervised learning
Part 2: Some introductory Methods for Regression
- Least Squares and Subset Selection
- Shrinkage Methods for Subset Selection (Ridge vs. LASSO)
Part 3: Some introductory Methods for Classification
- Introduction to classification and probability modeling
- Logistic regression (briefly)
- K-Nearest Neighbors
Part 4: Simple Decision Trees and how to improve their Predictive Power
- Decision Trees
- Random Forests
- Bagging, Boosting
Part 5: Model Selection
- Hyperparameters
- Bias-Variance trade-off
- Information Criteria
- Cross-Validation
Part 6: Support Vector Machines
Part 7: Neural Networks
- Single Hidden Layer
- Multilayer Perceptron
- Recurrent Neural Networks
Requirements:
Basic knowledge in linear algebra (especially matrix algebra), stochastics, geometry.
Basic knowledge in econometrics, especially linear regression, OLS, Maximum Likelihood.
Weitere Informationen zu den Prüfungsleistungen:
Assessment: Take - Home Assignment + Presentation
Literatur:
Hastie T., Tibshirani S., Friedman J., The Elements of Statistical Learning, Springer Series in Statistics, Second Edition (2009)
Abu-Mostafa Y., Magdon-Ismail M., Lin HT., Learning From Data, AMLBook, 2012
James G., Witten D., Hastie T., Tibshirani S., Introduction to Statistical Learning, Springer Texts in Statistics, First Edition (2013)
Wenn Sie E-Learning Funktionalitäten nutzen möchten, tragen Sie bitte "Ja" ein.:
"Ja"
|
