Sign In
CRM > English > Activities > Curs 2016-2017 > Reviewing core statistics
Reviewing core statistics
General Information

3, 10, 17 and 24 of November and 1 of December, 2016

Location: Room A1, at the Centre de Recerca Matemàtica

Schedule

List of participants

Course description

This activity is aimed at the researcher who has probably taken an introductory statistics and probability course at some stage and would like a brief introduction to the core methods of statistics and how they are applied. We are going to do a panoramic view by the good practices in scientific methodologies from statistical point of view.

The outline of course is over the Statistics as the base of the based on experience scientific method. The first ingredient are data and the second are the models. The third is how to quantify the link between them, called likelihood.

Table of contents:

 Part 1  - ​Datasets and Random Variables ​ ​ Data can have different nature and complexity, according to the way the are obtained and the way they are displayed (multidimansional data, functional data, dependent data,...). We shall review the process of data obtention and several ways to give them structure, according to their nature, their (possible) structures of dependence, based on the primigenial objective of the experiment that produced them. We shall review the probabilistic foundations of the main data reduction techniques and statistical analysis. Part 2  - ​Statistical Models and Probability theory ​In order to capture the essense of statistical modeling, we will begin with simple models. We then will add complexity in order to show a large amount of models, from an independent setting (parametric and non-parametric models) to learining models based on data mining techniques (neural network, random forest, ...) passing through classical regression and classification models (mixed model, generalized linear model, …) as well as temporal dependence (autoregressive, moving average,…). We will review these models from a probabilistic point of view. Part 3  - ​Likelihood and Inferential Questions ​Finally, we shall link the data with the models, this part includes model asessment, model validation and model selection. We shall see how the link will allow us to answer the inferential questions that gave birth to the experimental study (testing, predicting, forecasting, learning,...) and we will mention diferent approaches to give an answer from the statistical point of view (frequentist statistics, Bayesian statistics, data-based statistics).

Coordinators

Alvaro Corral, Centre de Recerca Matemàtica

Isabel Serra, ​Centre de Recerca Matemàtica

Organizing Commitee

Roberto de la Cruz, Centre de Recerca Matemàtica

Núria Folguera, Centre de Recerca Matemàtica

Marina Vegué, Universitat Politècnica de Catalunya

Lecturers

Alejandra Cabaña, Universitat Autònoma de Barcelona

Isabel Serra, Centre de Recerca Matemàtica

Registration

Registration fee: 50€

Deadline for registration: October 31, 2016

Contact information

If you have any questions please contact Alba Tomàs (atomas@crm.cat)