Blockkurs: Machine Learning and AI in TensorFlow and R

Location / Time: Zoom, 03.-07.05.2021,09:00-17:00

Contact: Prof. Dr. Florian Hartig (florian.hartig@ur.de)

Target group of the course: The main target group are MSc and PhD students with appropriate prior knowledge in R. BSc can join if they have the appropriate prior knowledge. Minimum requirement for this course is an introductory statistics lecture (e.g. Lecture 54134 Statistik und Bioinformatik), an introductory R course (e.g. 54 371, Blockkurs:Introduction to Statistics with R), or similar, and being able to comfortably operate in R / RStudio. In detail, these courses would teach you

If you don't feel comfortable with thes topics, block course 54 371 will be more suitable for you. Moreover, we recommend taking

before taking this course. If you haven't taken either of these course, we highly recommend reading "An Introduction to Statistical Learning" http://faculty.marshall.usc.edu/gareth-james/ISL/ before taking this course!

Learning objectives: This course provides a practical and general introduction into machine learning / predictive models with Google TensorFlow and Keras in R. We will cover the standard task in a practical data project, as well as todays's standard methods in machine learning and AI. 

Contents:

Course type / Language: Methods will be shortly explained, but the course will mostly concentrate on the practical aspects of running these algorithms in R. It is therefore highly recommended to participants to get some prior information on the methods (what the idea is, what they do) in advance of the course, either via studying the recommend readings, or through taking the recommended lectures on machine learning. 

Certificate / Evaluation: A pass of this course (unmarked) requires the successfull completion of a data analysis project after the course. PhD students that participate via RIGEL will be except from this, as they also receive 0.5 ECTS less (RIGEL regulations). 

LSF Link: https://lsf.uni-regensburg.de/qisserver/rds?state=verpublish&status=init&vmfile=no&publishid=170592&moduleCall=webInfo&publishConfFile=webInfo&publishSubDir=veranstaltung

 

Blockkurs: Fortgeschrittene Methoden der Biostatistik in R / Advanced Biostatistics

Location / Time: Zoom,12.-16.04.2021, 9:00-17:00

Contact: Prof. Dr. Florian Hartig (florian.hartig@ur.de)

Target group of the course:  Interested Students and PhD students of all disciplines. Minimum requirement for this course is an introductory statistics lecture (e.g. Lecture 54134 Statistik und Bioinformatik), an introductory R course (e.g. 54 371, Blockkurs: Introduction to Statistics with R), or similar, and being able to comfortably operate in R / RStudio. In detail, these courses would teach you

If you don't feel comfortable with thes topics, block course 54 371 will be more suitable for you. 

Learning objectives: The aim of this course is to introduce participants to modern statistical methods beyond what is typically covered in a first statistics class. We will cover commonly used advanced regression models (GLS, GAM, LMM, GLMM), resampling methods (cross-validation, bootstrapping), and some machine-learning applications, as well as guidance to deal with typical regression problems, such as heteroscedasticity or spatial autocorrelation.

Contents:

Methods

All methods will be taught with biological apllications, in particular

Course type / Language: Teaching will consist of short theoretical instructions, demonstrations in R, and exercises in R. Course language is English on request.

Certificate / Evaluation: PhD students can obtain 1.5 ECTS for attendance only (without project) according to RIGeL rules.

LSF Link: https://lsf.uni-regensburg.de/qisserver/rds?state=verpublish&status=init&vmfile=no&publishid=170584&moduleCall=webInfo&publishConfFile=webInfo&publishSubDir=veranstaltung

 

Introduction to Statistics in R

Location / Time: Zoom, 26.-30.04.2021, 9:30-16:30

Contact: Lisa Hülsmann (lisa.huelsmann@ur.de)

Target group of the course: Students and PhD students of biology and similar disciplines with basic statistical knowledge (e.g. compulsory bachelor lecture for biostatics) and with little or no prior knowledge in R.

Learning objectives: The aim of the course is to be able to understand and interpret fundamental concepts of statistics; to choose the right tool among basic statistical methods for typical (biological) studies; to apply these methods in the statistic software R; to interpret the results of these analyses correctly; and to design experiments and the collection of data.

Contents:

Basic statistics in R

  1. Types of data
  2. First steps in R
  3. Descriptive statistics and visualisation
  4. Hypothesis testing
  5. Linear regression
  6. Generalised linear models
  7. Basic multivariate statistics

Course type / Language: The course will be held as an online block course. Course language is English.

Certificate / Evaluation: PhD students can obtain 1.5 ECTS for attendance only (without project) according to RIGeL rules.

Software (in case you want to use your own laptop):
• R – freely available at https://www.r-project.org/
• RStudio – freely available at https://www.rstudio.com/

Course materials:
• at GRIPS

LSF Link: https://lsf.uni-regensburg.de/qisserver/rds?state=verpublish&status=init&vmfile=no&publishid=170588&moduleCall=webInfo&publishConfFile=webInfo&publishSubDir=veranstaltung

 

Vorlesung: Statistik und Bioinformatik (Wintersemester 2018/19)

Die Vorlesung führt in die grundlegenden Konzepte der Biostatistik und der Bioinformatik ein.
Den Schwerpunkt im Teil Bioinformatik bilden Verfahren zum Sequenzvergleich. Gleichzeitig werden theoretischen Grundlagen, grundlegende Konzepte der Informatik und wichtige Ergebnisse der Testtheorie vorgestellt.

Der Stoff des Biostatistikteils ist zusammengefasst im Skript “Grundlagen der Statistik”, verfügbar hier

In der letzten Vorlesungswoche am 5. und 7. Februar 2018 findet der letzte Teil (Versuchsplanung) der Veranstaltung "Design und Auswertung" statt.

Biostatistik: Einführung (Population, Stichprobe, Merkmale, Skalenarten), Wahrscheinlichkeitstheorie (Zufallsvariablen und ihre Verteilung, Spezielle Verteilungen), Schätzung unbekannter Parameter (Konfidenzintervalle), Formulieren und Prüfen von Hypothesen, Ausgewählte statistische Tests (t-Tests, Chi-Quadrat-Test), Korrelations- und Regressionsanalyse.

Bioinformatik: Sequenzvergleich, Scores und ihre Anwendung, phylogenetische Verfahren, Markov-Ketten und Modelle.

LSF Link: https://lsf.uni-regensburg.de/qisserver/rds?state=verpublish&status=init&vmfile=no&publishid=123381&moduleCall=webInfo&publishConfFile=webInfo&publishSubDir=veranstaltung