Table of Contents:
Rapid Analytics and Model Prototyping
The RAMP is a versatile management and software tool for connecting at the University of Paris-Saclay data science to domain sciences, which is the main mission of the CDS.
Similarly to a data challenge, the data provider arrives with a prediction problem and a corresponding data set. An experienced data scientist then cleans and curates the data, formalises the problem and sets up the problem using the RAMP software. When the data science problem requires the mastering of a specific tool, the RAMP event can be preceded by a Training Sprint. Part of the Training Sprint can also be devoted to introducing the domain science problem, otherwise this introduction takes place at the beginning of the RAMP.
The following RAMPS have been organised:
The IMaging-PsychAtry Challenge (IMPAC) is a data challenge on Autism Sprectrum Disorder (ASD). ASD is a severe psychiatric disorder that affects 1 in 166 children.
There is evidence that ASD is reflected in individuals brain networks and anatomy. Yet, it remains unclear how systematic these effects are, and how large is their predictive remain unclear. The large cohort assembled here can bring some answers. Predicting autism from brain imaging will provide biomarkers and shed some light on the mechanisms of the pathology.
More information: https://paris-saclay-cds.github.io/autism_challenge/
Mars crater detection
This challenge proposes to design the best algorithm using a collaboration strategy to detect crater position and size starting from the most complete Martian crater database containing 384 584 verified impact structures larger than one kilometer of diameter. We propose to give to the users a subset of this large dataset in order to test and calibrate their algorithm. We provide THEMIS nightime dataset, already projected to avoid distortion, sampled at various scales and positions in form 112×112 pixels images. Using an appropriate metric, we will compare the true solution to the estimation. The goal is to provide detection of more than 90% (crater center and diameter) with a minimum number of wrong detection.
Drug classification for Spectra
Chemotherapy is one of the most used treatment against cancer. To prevent wrong medication, some recent French regulations impose the verification of anti-cancer drugs before their administration. In this context, the goal of the Drug Classification RAMP was to develop prediction models able to identify and quantify chemotherapeutic agents from their Raman spectra.
The event took place on May 2016 at PROTO204.
The HiggsML RAMP was the first event of a series of bootcamps that the CDS was launching. This first session was about a gentle introduction to practical machine learning through a concrete application to the Higgs-ML challenge data (ATLAS experiment).
The event took place at Proto204 in january 2015 and we had a special guest, Gábor Melis, who recently won the Higgs-ML competition held by Kaggle.com.
The Health care RAMP took place on February 2015 at PROTO204 and it was the second edition of the CDS bootcamps in Machine Learning and Data Science.
Classifying variable stars
The Classification of variable stars RAMP took place at Proto204 on April 2015 and was on Astrophysics, more precisely, on classification of variable stars from their light curves (luminosity vs time profiles).
El Niño prediction
Similarly to the variable stars RAMP, in El Niño Prediction RAMP the pipeline consisted of a feature extractor and a predictor. This RAMP took place at PROTO204 on June 2015, and its objective was to predict six months ahead the temperature at surface (TAS) in the El Niño 3.4 region from TAS data simulated by the CCSM4 model.
The Pollenating Insects RAMP took place at PROTO204 on October 2015.
In this RAMP we classified images of pollenating insects from the SPIPOLL crowdsourcing project of the Paris Museum of Natural History (MNHN). The RAMP is brought to you by Romain Julliard (MNHN) and your regular coaches. We are grateful to the Université de Champagne-Ardenne ROMEO HPC Center and NVIDIA for providing the GPU backend and engineering support for the RAMP, and to Proto204 for hosting the event.
In the Macroeconomic Surrogate RAMP we learnt a surrogate model for an agent-based macroeconomic model (ABM) and an objective function. The goal was to have a fast filtering algorithm that can replace this slower simulation in, for example, a stochastic optimization or approximate Bayesian computation.
The event took place on February 2016 at the Maison des Sciences Économiques.
HEP detector anomalies
The purpose of the LHC Atlas RAMP was to detect anomalies in the LHC Atlas detector, to Separate a skewed data point from a original data point.
The event took place on May 2016 at the Auditorium Pierre Lehmann (LAL).