Methods for Statistical Inference

The general theme of the three-month event is to ask the questions: do we have the right methods and tools in place for the next decade in cosmology? What are the key challenges and promising new avenues to solve them? The "Statistical Inference" workshop will focus on the unique challenges of extracting cosmological information from data and new techniques that address them. Analytical and Computational/Numerical methods will be the focus of two other workshops.

This 5-day workshop will take place from the 22nd until the 26th of October 2018. The program will consist of 45-minute invited talks (three per half day, with Wednesday afternoon off and an early close on Friday) leaving plenty of time for discussion and collaboration. Some of these talks will be labelled "challenge talks"; those will focus on identifying the main open methodological problems that must be overcome to interpret existing and upcoming data sets in certain subfields of cosmology.

We have identified the following main topic areas for this workshop:

Statistical Computation

Hierarchical Models

Sparse Reconstruction/Compressed Sensing

Machine Learning

Likelihood-Free Inference

Monte Carlo Simulations

Semi-Blind Inference

Inverse methods

Systematics

Analysis on the Sphere

Machine-Aided Discovery

Programme:

To know the workshop programme, please click here

Scientific Organizing Committee:

Patrick Peter

Benjamin Wandelt

Matias Zaldarriaga

Practical informations:

October 22-26, 2018

Institut Henri Poincare - 11, rue Pierre et Marie Curie - Paris

For more information, please send an email to the following address: cosmomath18[at]ihp.fr

To go to the IHP website, click here