A number of tutorial sessions will be organized before the conference targeting primarily young researchers (at Master, PhD, and postdoc levels), as well as more senior researchers who are interested to learn new techniques. The goal of the tutorials is to present advanced data analysis methods (such as denoising, deconvolution, inpainting, detection, source separation, etc.), and to demonstrate how one may use the available codes, relevant to these new and exciting applications. These tutorials are not intended only for an audience with a background in astronomy.
Note: The Cargèse conference center can host 40 participants during the tutorial week.

Five tutorials are fixed (see below), and few others will selected from the submitted tutorials. A tutorial must be related to an open source software available on the web, and can be submitted via this link.

Tutorial 1: Wavelets, sparse representations and denoising
Organizers: Sandrine Pires (CEA Saclay)
Duration: 3h
This tutorial will give an introduction to the wavelet transform and other multiresolution transforms that provide a framework for decomposing images, volumes, and time-series into different forms that are more suitable for many applications. The wavelet transform has become a common tool for a variety of signal. The goal of this tutorial is to show that for some signals other representations such as ridgelet or curvelet transforms are more adapted. Many applications are within the scope of this course. During this tutorial, I will focus on the denoising application taking advantage of the sparse representation of the data. This course will be organized as practical works under IDL. Participants are encouraged to have their own laptops with IDL (7.0 or higher) installed.

Tutorial 2: iCosmo: CANCELLED

Tutorial 3: Inverse Problems (deconvolution) and Compressed Sensing
Organizers: Florent Sureau (CEA Saclay)
Duration: 3h
In this tutorial we will introduce the inverse problem of deconvolution and the compressed sensing approach and illustrate how sparsity of the looked-after signals is a crucial property for signal recovery. Deconvolution is a recurring problem in signal processing and has been extensively covered for the past 30 years. Such ill-posed problem (with an inverse operator potentially unbounded) require adequate regularization to limit noise amplification. In this tutorial, we will briefly review early deconvolution techniques still used in astronomy and illustrate how sparsity of the signals in appropriate dictionaries can be used to improve the deconvolution results. Compressed sensing is a new approach for data acquisition and for solving some inverse problems, relying on only a few incoherent measurements of sparse signals. In this case, the classical restoration problem without prior is underdetermined (no unique solution). In this tutorial, we will briefly present the theory of compressed sensing and illustrate how this approach has been used to acquire or process astrophysical datasets. In both cases, pratical exercices will be proposed to better illustrate the relative strength of the sparsity prior in these two situations.

Tutorial 4: Morphological Component Analysis to perform image decomposition and inpainting
Organizers: Sandrine Pires (CEA Saclay)
Duration: 3h
This tutorial will focus on the Morphological Component Analysis (MCA) of signals and images. This technique takes advantage of the sparse representation of the data in large overcomplete dictionaries. During this tutorial, the concept of sparsity and morphological diversity will be described and I will review two major applications based on morphological diversity: image decomposition and inpainting. abstab During this tutorial, I will present some applications to astrophysical data such as CMB data, Weak Lensing data or Corot data. This course will be organized as practical works under IDL. Participants are encouraged to have their own laptops with IDL (7.0 or higher) installed.

Tutorial 5: Astroinformatics - The New Paradigm of Astronomy in the Epoch of Petabyte-Scaled Archives
Organizers: Petr Skoda (Astronomical Institute Ondrejov, Academy of Sciences, Czech Republic)
Duration: 3h
Astroinformatics is a new emerging discipline of astronomy, promising to extract a new scientific knowledge from the contemporary deluge of astronomical data. As the data have been growing faster than a computer technology can cope with, a qualitatively new research methodology is required, based on advanced statistics and data mining methods as well as on a new approach to sharing huge databases in a seamless way by the global research communities. The indispensable part of this so-called e-Science paradigm is represented by the project of astronomical Virtual Observatory, the global infrastructure of federated astronomical archives, web-based services and powerful client tools supported by supercomputer grids and computing clusters. In the first part of the tutorial we give a overview of the motivations, early history and technological principles of Virtual Observatory (VO) as well as a more philosophical view of data mining, Citizen Science and new promises of informatics in the age of data-intensive science. In the second part the practical usage of some commonly used VO tools and services will be demonstrated on solution of simple problems. Participants wanting to follow the practical exercises on their own computers are required to install the Oracle (Sun) Java 1.6 or later with Webstart support.

Tutorial 6: Data Processing on the sphere with applications to Astrophysics
Organizers: Florent Sureau (CEA Saclay)
Duration: 3h
In this tutorial, we will familiarize the participants with the particular issues and tools involved in analyzing data acquired on the sphere. Such situation is commonly met in modern astrophysics, for instance in full-sky experiments such as WMAP or Planck, and requires specific analysis tools such as the MRS package. We will present different sampling schemes, focusing on Healpix, introduce the spherical harmonics expansion and various multiscale decompositions such as undecimated isotropic wavelet transform on the sphere or curvelet on the sphere. These fundamental tools will then be employed in practical restoration problems such as denoising or inpainting of WMAP data.This course will be organized as a series of practical works under IDL. Participants are encouraged to have their own laptop, with IDL (7.0 or higher) installed, as well as the Healpix and MRS packages.

Tutorial 7: Reflex: Scientific workflows for the ESO pipelines with a focus on UVES and XSHOOTER
Organizers: Dan Bramich & Wolfram Freudling (ESO)
Duration: Morning or afternoon session
Reflex is a workflow based environment to run ESO VLT pipelines. It includes interactive tools to inspect intermediate results and change recipe parameters. Furthermore, Reflex workflows provide the important capabilities of automatic data organisation, visual representation of the data flow, interactive windows, and the potential for user modifications. This tutorial will provide an introduction to using Reflex. We will cover the details of how to use and interact with the workflows to obtain the best data reduction while working specifically with the workflows that have been developed for the reduction of UVES and XSHOOTER data from the VLT. Hence this tutorial will also be of interest to those users that would like to optimally reduce their UVES or XSHOOTER data with the least hassle. We will finish the tutorial with a look at how workflows can be modified and created from scratch.

Tutorial 8: Blind source separation methods: application in cosmology
Organizers: Jerôme Bobin (CEA Saclay)
Duration: 3h
This tutorial will give essential insights into the use of blind source separation methods in astrophysics with a particular application in cosmology. Multichannel or multiwavelengths data are more than common in the field of astronomy and astrophysics. Analyzing this particular multidimensional data requires specific tools; these data are commonly modeled as the superposition of elementary components or sources which generally do not contribute similarly in each observation channel. The aim of blind source separation is to estimate simultaneously the sources and their contribution weight in each of these channels. Think for instance of a symphonic concert recorded with different microphones (i.e. the observation channels) and imagine the problem of separating out the contribution of each individual instrument from this cacophony. This tutorial will review the basics of blind source separation as well as more recently introduced sparsity-bases techniques. An application to WMAP data will also be at the menu. This course will be organized as practical works under Matlab. Participants are encouraged to have their own laptops with Matlab (7 or higher).