Workshop R Programming and Mixtures

Presentation:

This one-day course, developed by Dr Jochen Einbeck (Durham University, UK) and Professor John Hinde (NUI Galway, Ireland), consists of two parts, where the first part refreshes generic R programming skills, while the goal of the second part is the practical implementation of inferential tools for finite Gaussian mixture modelling.

The workshop assumes basic working knowledge with R, though it does not require advanced programming skills. More specifically, the workshop begins with recalling basic tools and concepts which are useful for R programming in general, these include: workspace handling, reading in data files, extracting information from vectors and matrices, basic operations with data frames (such as ordering); basic programming skills such as if/then, while, for, and apply, and the construction of functions.

In a short lecture to the beginning of the second part, the workshop turns to Finite Gaussian Mixture models. The idea of 'complete data' (assuming the unknown component memberships to be known) is explained, based on which the complete likelihood is constructed and maximized. It is demonstrated how the resulting estimators can be incorporated into an EM algorithm, which is used to estimate all parameters of the mixture model. Using skills acquired in the first part of the workshop, the EM algorithm is implemented in R in the practical part of the afternoon session (which requires implementation of the E--step, the M--step, and the EM wrapper function). Depending on the progress, the more advanced students can proceed with implementing a bootstrap test for the number of mixture components. The techniques are illustrated with, and applied to, real data sets taken from astronomy and the energy sector.

Professor:

Jochen Einbeck - Senior Lecturer in Statistics, Department of Mathematical Sciencies, Durham University

Audience:

The Workshop R Programming and Mixtures is addressed to professional or academic statisticians and researchers who are interested in being introduced to the use of these methods.

It is expected that the participants are familiar with R software.

Those attending this session must bring their computer with the software R.

Organization details:

The Workshop R Programming and Mixtures will be held on December 12th, 2014, from 9:30 to 17:00.

The workshop duration is 5.5 hours.

Limited places.

Pre-registration will be formalized via the Servei d'Estadística Aplicada filling out the registration form that you will find on the web. Once we received your form we will send an email to confirm that either you fave an assigned place or you are on the waiting list.

Registration fees (2014):

Concept Quantity Import
    External Mixed UAB
Registration 
(before November 30th)
1 assist 160,00 €  160,00 €  95,00 €
Registration
(afer November 30th)
1 assist 205,00 €  205,00 €  135,00 €

UAB rate: People belonging to the UAB university community may benefit from this rate (PAS, teachers, students) as well as students from other universities as long as they send a copy of the current course registration. In case an invoice is required they will have to be registered with another fee.

Grants for Degree of statistic students, see conditions on the registration form.

Payment details:

Once the pre-registration is completed you will receive an email informing of the details for the registration payment.

People interested in applying for an invoice in the name of a company must state in the payment proof the name of their organization and NOT of the attendant of the course himself. Once the payment of the course has been made there will be no refund unless there are circumstances beyond one’s control.

Do wait for our confirmation of the reservation for the course before payment.

Course program:

First session: Basic R programming

  • Preliminaries: Working directory, data frames, and workspace
  • Basic programming and operations with data frames

Lecture session

  • Motivation and theory for finite Gaussian mixtures

Second practical session:

  • Implementing the EM algorithm for mixture models
  • Simulation from Gaussian mixtures
  • Advanced and challenging...

 

 

Campus d'excel·lència internacional U A B