Data Science @Saclay

The official account of the Paris-Saclay Center for Data Science, an interdisciplinary project around data.

[Data Science @Saclay @SaclayCDS]

The new https://t.co/yEwOA8vBc7 data challenge platform by @SaclayCDS is out and kindly hosted by @OVHcloud_FR ! thanks a lot #DataScience

[Data Science @Saclay @SaclayCDS]

. @SaclayCDS
hosted another very successful workshop on Introduction to Data Science in Python. Find out more: https://t.co/LulsHbRtbp
Stay tuned for the workshops coming in 2020

[Balázs Kégl @balazskegl]

Can #AI learn #Physics? #DataScience challenge by @HuaweiFr with prize pool. Forecast the behavior of a dynamical system. Open to French students. Number of registrations is limited: https://t.co/1T0RkvBBID. Please share.

[Data Science @Saclay @SaclayCDS]

The @SaclayCDS recently organised an introduction to data science workshop. Find more about this event and looking forward to the next session: https://t.co/Zkvnz3JUB2

[Balázs Kégl @balazskegl]

Workshop March 7 on setting up your own RAMP server 1) for open code submission #datachallenges 2) for teaching #datascience or 3) for managing your datascience team. Please sign up here https://t.co/eOmks5QKQV
@SaclayCDS @jorisvdbossche @glemaitre58 @agramfort @UnivParisSaclay

[Guillaume Lemaitre @glemaitre58]

The Paris-Saclay CDS organizes a
workshop "Introduction to Data Science in Python" March 13-14
at Inria Saclay Ile-de-France. The event is opened to all
scientists of Paris-Saclay area. Register to https://t.co/ocEk96eoIa

[Balázs Kégl @balazskegl]

Our next single-day @SaclayCDS #RAMP hackathon will be on October 10 at the INRIA Turing building on detecting Solar storms from spacecraft measurements (event detection in time series).

If you would like to participate, please sign up before October 3:
https://t.co/oSyYzfFr7H

[Balázs Kégl @balazskegl]

We have concluded the #ASDchallenge of @SaclayCDS @institutpasteur @IESFfrance, 123 participants, 677 submissions, AUC moved from 0.68 to 0.81. Congrats to the winners and welcome to the collaborative phase! #datascience for #healthcare #MachineLearning
https://t.co/CkMxxRTsBH

[trackml @trackmllhc]

Edwin Steiner lands at second place 75.6% in our #trackml #kaggle competition. outrunner still leading at 80.7%. uval r third at 72.8%. Here a viz showing the evolution of the score of 100 best teams. #hepml

[Sebastien Treguer @ST4Good]

Machine Learning bootcamp for TrackML challenge on @CERN data https://t.co/xPkIGhjw1N @LaPaillasse Paris June 18th #deeplearning #openscience

[trackml @trackmllhc]

This will be a day long boot camp. Please consider coming if you are in Paris next Monday (registration mandatory). https://t.co/yk99YfHuJf

[trackml @trackmllhc]

Zidmie just snatches first place to Mickey who had been holding it for 10 days, 64.9% vs 64.1% in our #trackml #kaggle challenge. Kudos to Grzegorz Sionkowski posting a 43% kernel (still DBScan). Guys, we gave you DBScan to get started, don't get locked into it !

[trackml @trackmllhc]

Interesting to see @hengcherkeng, one of the most active contributor to the forum of our #trackml competition (and overall #kaggle discussion grandmaster) inching step by step, now number 4. Also good to see the pack getting closer to the lead.

[trackml @trackmllhc]

A nice vis from @hengcherkeng in our #trackml @kaggle challenge. This is a small portion of an event. Grey the hits and ground truth tracks.
Red: predicted pairwise link. Coordinates are x y z in millimeters. https://t.co/gsN97eGoT2

[trackml @trackmllhc]

Congrats to Vicens Gaitan @ControlByte snatching leadership from Zidmie (65.5% against 64.9%) in our #trackml @kaggle competition. Mickey still third with 64.1%. 3 more above 60%. Good to see 24 more above 50%. #hepml .

[trackml @trackmllhc]

Yep, this is us! 3 years in the making! https://t.co/besRPzYA0e

[Alexandre Boucaud @alxbcd]

Preprint on our #RAMP #datascience platform available now on openreview. @SaclayCDS https://t.co/xgMn2UAvEP

[Gael Varoquaux 🦋 @GaelVaroquaux]

New paper: Similarity encoding for learning with dirty categorical variables
https://t.co/j2JQxy5FcC

Code on https://t.co/9Saecz75MP

Useful for data scientists fighting with dirty data

[Paragios, Nikos @AgoniGrammi]

Can #medicalimaging #ArtificialIntelligence predict treatment success of #Immunotherapy
Our results @ASCO #ASCO18 indicate strong correlations
https://t.co/nAr23E6Z4A
#PhD of @rgrsun
@GustaveRoussy,@Inserm,@centralesupelec,@u_psud,@ParisSaclay,@Inria with @charles_ferte,@Edradonc

[Data Science Game @datasciencegame]

This is THE day ! The qualifying phase has just begun today 👩‍💻👨‍💻
Latecomers : you still have time to register, but hurry up since the game is on ! 👀
Do your best... and you may be one of the 20 finalist teams !🏆