Description: I

Within this seminar knowledge is imparted about the basics of medical statistics and epidemiology. An introduction to descriptive statistics is given as well as to the construction and interpretation of a confidence interval. The principle of statistical testing will be explained and basic topics on epidemiology will be discussed. The concept of correlation and linear regression is taught. Important statistical aspects that play a role within the planning of a confirmatory clinical study are introduced. The principle of meta-analysis will also be explained.

Sessions (do not build on one another):

• Descriptive statistics and confidence intervals
• Statistical testing
• Introduction to epidemiology
• Correlation and linear regression
• Planning of a confirmatory clinical study
• Introduction to meta-analysis

Languages: German and English (alternating)

Scope:

• 6x2 teaching units

Next Dates:
TBA

Location:
TBA

Registration: Via Moodle or by email to phd-info@uke.de

Contact:

Ann-Kathrin Ozga a.ozga@uke.de

Prodekanat für Forschung phd-info@uke.de

Description:

In this lecture series you will learn about event-time analysis and basic principles of machine learning. The introduction to diagnostic studies is also part of the lecture series. The handling of clustered data and missing values will be presented.

Prerequisite: Basic knowledge in medical statistics (see e.g. our lecture series Introduction to Medical Statistics and Epidemiology)

Sessions:

• My rate to die - Time to event analysis
• Unsupervised machine learning or what machines can learn from unlabelled data
• Supervised machine learning and ensemble methods - the more the merrier?
• How to evaluate screening methods - Diagnostic studies
• Cluster in your data - Mixed models (hierarchical or multilevel models)
• Fill the gap - How to handle missing data

Language: English

Scope:

• 4x3 teaching units

Next Dates:

TBA

Location:

TBA

Registration: Via Moodle or by email to phd-info@uke.de

Contact:

Ann-Kathrin Ozga, a.ozga@uke.de or
Prodekanat für Forschung phd-info@uke.de