diiP Summer School 2024 - dSDS

10 juin 2024, 08:00 → 14 juin 2024, 18:30 Europe/Paris

418C, Halle aux Farines (Paris)

Themis Palpanas (LIPADE - Paris Descartes University), Yvonne Becherini (Université Paris Cité)

 

The Data Intelligence Institute of Paris (diiP) is an interdisciplinary initiative of Université Paris Cité. It is a laboratory that fosters and supports the emergence of interdisciplinary practices around data science and data intelligence. It gathers researchers from formal sciences, physical sciences, life sciences and social sciences.

The diiP organizes the 1st diiP Interdisciplinary Summer School on Data Science (dSDS), which will focus on Deep Learning Data Analytics Techniques, and will involve a set of invited keynote speakers and lecturers on these subjects. The dSDS will take place in Paris (France), at the Université Paris Cité, campus Grands Moulins.

The dSDS is addressed to early-stage researchers, such as graduate (MSc and PhD) students, postdocs, early-career faculty, and other academic and industrial researchers interested in the diverse field of Data Science. Advanced undergraduate students with the appropriate background will be considered as well.

The organization of the school is sponsored by diiP. This sponsorship covers all lectures, accommodation for participants living far from Paris (4 nights), coffee breaks and lunches, social activities and a social dinner. Any other expenses will be the responsibility of the participants and of their institutions.

All interested participants need to apply through the Registration form by uploading a 1-page CV. Applications are received until April 15, 2024. 

Applications will be reviewed, and accepted participants will be contacted for a confirmation of their participation.

lundi 10 juin

08:00 Travel to venue

13:00 Lunch

1 AI for Research

Orateurs: Themis Palpanas (LIPADE - Paris Descartes University), Yvonne Becherini (Université Paris Cité)

diiP School-1.pdf

16:00 Coffee break

2 Large Language Models (LLMs)

Large language models have become ubiquitous. This lecture introduces to language modelling and large language models. We will discover what they are, where they come from and the primary motivations behind their design. We then provide an overview of the properties of these models when trained at the current scale of very large language models.
If time remains, we introduce the problematic of explaining their behaviour.

Orateur: Prof. Benoît Crabbé (Université Paris Cité)

llms_BC.pdf

18:00 Networking/Free time

19:00 Social Dinner - Bouillon Racine

mardi 11 juin

3 Supervised Learning

Orateur: Prof. Alexandros Iosifidis (Aarhus University)

Lecture_Supervised_Classification_diiP2024.pdf

10:30 Coffee break

4 Supervised Learning, Hands-on session

Orateur: Prof. Alexandros Iosifidis (Aarhus University)

5 AI in Medicine/Biology

Orateur: Prof. Guillaume Assié (Institut Cochin, Inserm CNRS Université de Paris Cité)

240611_IAmedicine_diipSummerSchool.pdf

13:00 Lunch

6 Representation Learning

Orateur: Dr Jenni Raitoharju (University of Jyväskylä)

representation learning2.pdf

16:00 Coffee break

7 Representation Learning - Hands-on session

Orateur: Dr Jenni Raitoharju (University of Jyväskylä)

17:30 Networking/Free time

mercredi 12 juin

8 Knowledge-guided Data Science

This lecture presents an overview of knowledge-guided data science, a rising methodology in machine learning which fuses data with domain knowledge. We will present numerous case studies on this methodology to showcase how to unleash its potential in real-world data science applications.

Orateur: Shen Liang

20240612_diipSummerSchool_KnowledgeGuidedDataScience.pdf

10:30 Coffee break

9 Knowledge-guided Data Science - Hands-on session (Basic)

Orateur: Shen Liang

ShenLiang-HandsonSession.pdf

13:00 Lunch

10 High-Dimensional Vector Similarity Search

Very large amounts of high-dimensional data are now omnipresent
(ranging from traditional multidimensional data to time series and deep
embeddings), and the performance requirements (i.e., response-time and
accuracy) of a variety of applications that need to process and analyze
these data have become very stringent and demanding. In the past years, high-dimensional similarity search has been studied in its many
flavors. Similarity search algorithms for exact and approximate, one-off
and progressive query answering. Approximate algorithms with and without
(deterministic or probabilistic) quality guarantees. Solutions for
on-disk and in-memory data, static and streaming data. Approaches based
on multidimensional space-partitioning and metric trees, random
projections and locality-sensitive hashing (LSH), product quantization
(PQ) and inverted files, k-nearest neighbor graphs and optimized linear
scans. Surprisingly, the work on data-series (or time-series) similarity
search has recently been shown to achieve the state-of-the-art
performance for several variations of the problem, on both time-series
and general high-dimensional vector data. In this talk, we will touch
upon the different aspects of this interesting story, and present some of the state-of-the-art solutions.

Orateur: Themis Palpanas (LIPADE - Paris Descartes University)

Palpanas-ADS-20.pdf

16:00 Networking/Free time

19:15 Seine River Cruise and Social Dinner

jeudi 13 juin

11 An Overview of Anomaly Detection for Time Series

Anomaly detection is an important problem in data analytics with applications in many domains. In recent years, there has been an increasing interest in anomaly detection tasks applied to time series. In this talk, we take a holistic view of anomaly detection in time series, starting from the core definitions and taxonomies related to time series and anomaly types, to an extensive description of the anomaly detection methods proposed by different communities in the literature. We will then present new benchmarks capturing diverse domains and applications for the purpose of evaluating anomaly detection methods. We will then conclude on Ensembling and Model Selection for time series anomaly detection, discussing different strategies applicable to automatically selecting the appropriate methods for a specific time series.

Orateur: Dr Paul Boniol (INRIA & ENS)

Lecture_TSAD.pdf

10:30 Coffee break

12 Time Series Anomaly Detection in Practice - Hands-on session

Orateur: Dr Paul Boniol

tutorial repository

13 AI in Industry

Orateur: Dr Paul Poncet (Engie)

ENGIE_Data_Science_AI_diiP_2024-06.pdf

13:00 Lunch

14 Graph Neural Networks

Orateur: Prof. Michalis Vazirgiannis

DIIP_2024_GNNs.pdf

16:00 Coffee break

15 Graph Generative AI + Applications

Orateur: Prof. Michalis Vazirgiannis (École polytechnique, Palaiseau)

2024_06_generative_AI for graphs_uparistutorial.pdf

18:00 Networking/Free time

vendredi 14 juin

16 Generative adversarial networks and Active Learning

Orateur: Prof. Amal Saadallah (TU Dortmund University)

summer-school-diip.pdf

10:30 Coffee break

17 Generative adversarial networks and Active Learning - Hands-on session

Orateur: Prof. Amal Saadallah (TU Dortmund University)

hands-on-pres.pdf

18 AI in Particle Physics

Particle physics deals with gigantic machines, large quantities of experimental data and computer simulations, complex and lengthy theoretical calculations. It is the perfect playground to take advantage of machine learning algorithms. After a short introduction to high energy physics, this lecture will show how one can speed up steps like event generation or detector simulation, better measure parameters, improve classification of events as signal or background, or discover anomalies in data taking, through various applications of machine learning in this field.

Orateur: Dr Yann Coadou (CPPM, Aix-Marseille Université, CNRS/IN2P3)

diiPSummerSchool_AIinParticlePhysics240614.pdf

13:00 Lunch

14:30 End of the diiP School