- Boniol, P., and Palpanas, T. Series2graph: Graph-based subsequence anomaly for detection time series . Proc. VLDB Endow. 13, 12 (July 2020), 1821–1834.
- Schneider, J., Wenig, P., and Papenbrock, T. Distributed detection of sequential anomalies in univariate time series . The VLDB Journal 30, 4 , 579–602.
- Tiano, D., Bonifati, A., and Ng, R. Featts: Feature-based time series clustering . In Proceedings of the 2021 International Conference on Management of Data (New York, NY, USA, 2021), SIGMOD '21, Association for Computing Machinery, p. 2784–2788.
- Heng, Z., Yang, Y., Jiang, S., Hu, W., Ying, Z., Chai, Z., and Wang, C. Time2graph+: Bridging time series and graph representation learning via multiple attentions . IEEE Transactions on Knowledge and Data Engineering , 1–1.
- Boniol, P., Panagopoulos, G., Xypolopoulos, C., Hamdani, R. E., Amariles, D. R., and Vazirgiannis, M. Performance in the courtroom: Automated processing and visualization of appeal court decisions in France. NLLP workshop of the KDD Conference .
- Can a unique graph embedding method be more accurate on multiple analytical tasks than specific methods for each task?
- Can a unique graph embedding be constructed for large heterogeneous multivariate time series that preserves accuracy while remaining scalable?
- How can such embedding be used to interpret and explain analytical tasks (e.g., classification, clustering, anomaly detection)?
- Acquire an exhaustive understanding of the literature on graph representation for time series and graph-based methods for time series analytics.
- Propose and implement a new graph representation for multivariate time series.
- Evaluate the proposed solution on publicly available benchmarks (UCR-Archive for classification and clustering and equivalents of TSB-UAD for multivariate time series).
- Study the impact of heterogeneous (i.e., various acquisition rates, types of time series, stationarity, etc.) multivariate time seires on unified graph embedding.
- Propose interpretable solutions based on the graph representation of time series for specific analytical tasks.
- Write scientific research papers with the objective to publish them on top data analytics and data management conferences and journals.
- Subsidized meals
- Partial reimbursement of public transport costs
- Leave: 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.)
- Possibility of teleworking
- Flexible organization of working hours (after 12 months of employment)
- Professional equipment available (videoconferencing, loan of computer equipment, etc.)
- Social, cultural and sports events and activities
- Access to vocational training
-
Responsable Webmarketing
il y a 6 jours
Data Recrutement Paris, FranceOffre publiée le Paris · - Fonction Traffic manager seo sea · - Fonction Brand manager · - Fonction Social media manager · - Fonction Content manager · - Taille entreprise de 21 à 50 · - Télétravail ponctuel · - Technologies Adobe · - Technologies Google analytics · - Expérience ...
-
Business Developer Agence Senior
il y a 1 semaine
digiRocks Paris, France**Rejoins 250 experts acquisition agence**: SEO, SEA, SMA, Display, Social & Influence. · Benjamin recrute un(e) Business Developer Agence à Paris en CDI 70k€ + 30k€ · Expérience 5 ans - Expertise Agence SEO et/ou SEA et/ou Analytics - Télétravail 1-2 jours / semaine · **MISSION* ...
-
Associate/economist (Climate), Economic Policy
il y a 1 semaine
World Bank Group Paris, France**Job #**: · - req21480**Organization**: · - IFC**Sector**: · - Economics**Grade**: · - GF**Term Duration**: · - 3 years 0 months**Recruitment Type**: · - Local Recruitment**Location**: · - Paris,France**Required Language(s)**: · - English**Preferred Language(s)**: · - Spanish, F ...
-
Associate/economist (Climate), Economic Policy
il y a 1 semaine
IFC Systems Corporation Paris, France**Associate/Economist (Climate), Economic Policy Research** · **Job #**: · - req21480**Organization**: · - IFC**Sector**: · - Economics**Grade**: · - GF**Term Duration**: · - 3 years 0 months**Recruitment Type**: · - Local Recruitment**Location**: · - Paris,France**Required Langu ...
-
Front-end Engineer
il y a 1 semaine
Gorgias Paris, FranceEverything we do is for our customers, and we're currently serving over 12,000+ ecommerce merchants, including : Steve Madden, Timbuk2, Decathlon, and Sports Illustrated. They love us for our innovative product, our focus on their ecommerce needs, and, of course, our lightning-fa ...
-
Webmaster (F/H)
il y a 6 jours
BAYARD SA Paris er, FranceDescriptif du poste · **CDD jusqu'au 16 septembre 2024** · Vos principales missions seront les suivantes: · - L'animation, la mise à jour des sites des revues Études et Christus, l'intégration des articles print vers le web (exports xml), le lien avec le prestataire de développem ...
-
Stage M2
il y a 2 semaines
ADDACTIS Group Paris e, France**Rejoignez Addactis pour votre stage ou alternance et intégrez notre programme « Jeunes Talents » ** · **En tant que jeune actuaire en formation, vous souhaitez probablement que vos premières expériences en stage ou alternance vous permettent non seulement de mieux découvrir l'é ...
-
Data Scientist
il y a 1 semaine
Artefact Paris, FranceOver the last few years, Data Science at Artefact Benelux has skyrocketed in terms of novel projects coupled with a growing diverse international team. We tackle use cases that bring **Business value** and serve different countries. Our area of expertise involves **Applied Mathem ...
-
Stagiaire en Growth Marketing
il y a 3 semaines
HelloFresh Paris, France**Le poste**: · Nous sommes à la recherche d'un.e stagiaire Growth Marketing rockstar au sein de notre équipe Marketing. En tant que "Growth Marketing Operations Intern", tu seras au cœur de notre croissance, en développant de nouvelles campagnes pour faire découvrir HelloFresh à ...
-
Balenciaga - Rtw Sales Merchandiser (H/F)
il y a 6 jours
Groupe Kering Paris, FranceSummary · About us · Cristóbal Balenciaga founded the House in 1917 in his home of Spain. In 1937, he established the brand in Paris, designing its collections there until 1968. Cristóbal Balenciaga had a reputation as a couturier of uncompromising standards and was referred to a ...
-
Data Analyst
il y a 2 semaines
Mirakl Paris, FranceEnterprise marketplaces are growing at more than twice the rate of overall eCommerce. Mirakl's mission is simple. We help our 350+ clients seize this opportunity by providing the industry's first and most advanced enterprise marketplace SaaS platform, unparalleled expertise, and ...
-
Data Engineer
il y a 2 semaines
Data Recrutement Paris, FranceOffre publiée le Paris · - Fonction Data engineer hadoop spark · - Taille entreprise de 51 à 200 · - Télétravail partiel · - Technologies Data studio · - Technologies Gcp · - Technologies Python · - Technologies Sql · - Expérience 3 à 5 ans · - Expérience 6 à 10 ans · - Statut CD ...
-
Product Designer Senior
il y a 2 semaines
Mentorshow Paris e, France**About us** · At MentorShow, we strongly believe that online training has the power to unleash your creativity, to soften your stress, to introduce new passions into your life. · Most people don't have access to high quality, well crafted, in-person classes. With MentorShow, you ...
-
Growth Marketing Specialist
il y a 1 semaine
Fixter Ltd Paris, France StageSHIPCompany Description · We founded Fixter to bring car maintenance into the 21st century. Because we believe booking a car service or repair in this day and age should be as simple as booking a taxi, or ordering a takeaway. Simple, streamlined and stress-free. And completely on dem ...
-
Insights Analyst Intern
il y a 1 semaine
Teads Paris, France**About the role**: · We are looking for an** Insights Intern** to join our team and work across research projects, providing insight for advertisers and media agencies and support the sales team. · The role will involve providing support to the Research team, including setting u ...
-
Senior Quantitative Researcher Equities
il y a 2 semaines
S.R Investment Partners Paris, France-S.R Investment Partners · Paris, France · Posted 1 hour ago Permanent Competitive + Bonus · - POSTED BY · - Margarita Ivlieva · - RecruiterFollow · - A renowned Hedge Fund is looking for a Senior hands-on Quantitative Researcher in Equities to lead a research in the main and the ...
-
Accounting Assistant
il y a 2 semaines
Dioxycle Paris, France**About Dioxycle** · Dioxycle is pioneering breakthrough carbon utilization technologies that convert industrial emissions into sustainable chemicals with unprecedented energy and cost efficiency. By displacing fossil fuels for the production of key chemicals, Dioxycle has the po ...
-
Customer Relationship Manager B2B
il y a 2 semaines
Lomographische GmbH Paris, FranceResponsable relation clients B2B · **LIEU**: Siège de Lomography à Vienne, Autriche / Paris · Avez-vous un talent pour partager vos passions et aimez-vous le challenge? Avez-vous l'esprit d'un entrepreneur et mettez-vous tout en œuvre pour atteindre et dépasser les objectifs qui ...
-
Senior Analytics Engineer
il y a 1 semaine
Algolia Paris, FranceAt Algolia, we are passionate about helping developers & product teams connect their users with what matters most in milliseconds · **OUR TEAM** · We're on a mission to make Algolia a data-driven organization, and we're looking for a Senior Analytics Engineer to join our Data & A ...
-
Insight Analyst
il y a 1 semaine
Mirakl Paris, FranceMirakl is a global SaaS technology company that enables businesses to achieve profitable and sustainable eCommerce growth. · Mirakl's industry-leading suite includes solutions in marketplace, dropship, supplier catalog management and pay-out, supplier sourcing ecosystem, personal ...
PhD Position F/M Campagne Doctorant Graph Representation for Multivariate Time Series Analytics - Paris, France - INRIA
Description
Contexte et atouts du poste
Massive collections of time-varying data (i.e., time series or data series in general) are becoming a reality in virtually every scientific and social domain. Examples of fields that involve data series include finance, environmental sciences, astrophysics, neuroscience, engineering, and multimedia. What is challenging about these data is that they are mainly highly multivariate, and the different dimensions that compose them may originate from different sources.
However, this high number of dimensions from different sources causes severe limitations. First, existing solutions employ one model per dimension or data type. This implies (i) a drop in accuracy because of missed correlations among important dimensions, (ii) a significant increase in execution time because of all the independent models that are used, and (iii) a drop in interpretability because of the multitude of embedding produced by all independent models. To reach efficient and scalable analysis without sacrificing accuracy, we need a unified data embedding that can enable multiple analytic tasks (such as anomaly detection, classification, and clustering) on multivariate and heterogeneous data series.
The objective is to move towards a unified data embedding that allows multiple analytic tasks (such as anomaly detection, classification, and clustering) on multivariate and heterogeneous data. Towards that direction, we proposed in past research Series2graph, a method that summarizes univariate time series into a graph [1,2]. Even though the latter method has been proposed mainly for anomaly detection, similar graph embedding for time series has demonstrated state-of-the-art and scalable results for tasks such as clustering [3], classification, and representation learning [4]. The benefit of such time series graph representation is three-fold. (i) First, such graph representation is easy to interpret by any user. (ii) Second, it can benefit from other graph-represented data (such as ontologies and knowledge graphs and textual data represented as graphs [5]). (iii) Last, one unified embedding can significantly reduce the analysis execution time (as shown for anomaly detection [1]).
However, no method exists that proposes a unified graph embedding for multivariate time series. The straightforward solution would be to build one graph embedding per dimension and then analyze them all together. However, the graph size would be linearly proportional to the number of dimensions, making it impossible to use in practice. In the case of heterogeneous multivariate time series, no holistic graph representation exists, and we need novel approaches to address this problem.
References:
Supervision:
The thesis will be co-supervised by Paul Boniol (Valda team, DI ENS & Inria Paris) and Michaël Thomazo (Valda team, DI ENS & Inria Paris). The PhD student will be part of the Valda team within the Computer Science Department of the École normale supérieure. Registration for the thesis will be carried out at Université PSL, via École doctorale 386 (Sciences mathématiques de Paris Centre). The doctoral student will benefit from the environment and resources of the VALDA team, the DIENS, the Inria Paris Center, and the PRAIRIE Institute, including local computing clusters. In addition, the PhD student will have access to the IDRIS Jean Zay supercomputer for GPU-intensive tasks.
Mission confiée
Research objective
The objective of this Ph.D. is to propose new meaningful graph representation and transformation for multivariate time series that could support basic analytics (classification, clustering, and anomaly detection). Overall, the research questions tackled are the following:
Application and use cases:
Time series analysis is a very important task for electricity production relevant applications. Indeed, the desire to analyze a large quantity of data efficiently and be able to express complex queries (i.e., anomalies discovery) can be crucial for industrial actors like EDF. For instance, one crucial goal for EDF is to improve the safety and availability of its electrical power plants by detecting anomalies that could occur. As massive gains are expected from reducing maintenance volumes, there is thus a serious need to have accurate and efficient algorithms to detect anomalies and understand their origins. Moreover, EDF has collected sensor data in every nuclear power plant for decades (at least 20 years). With a total of 58 nuclear power plants and more than 2000 sensors per unit, it represents a database of approximately 500 TeraBytes. Considering that the electrical power plants were built 20 to 30 years ago, we can expect that recent maintenance and new power plants will see their number of sensors and acquisition rate significantly increase, resulting in an exponential increase of new data. Moreover, half of the 2000 EDF electrical power plant sensors are boolean sensors, and the remaining half measure either water flow, pressure, temperature, or water level from very different parts of the plant, making each sensor almost unique. In addition to these already highly heterogeneous data series, the EDF database contains multiple logs (i.e., textual data) and structural knowledge (i.e., knowledge graphs representing the structure of the plant and the link between sensors). The context described above is highly related to the problems that will be tackled in this PhD. Thus, benefiting from previous collaborations of Paul Boniol with the research department of EDF, the PhD candidates might have the opportunity to apply the research conducted on such use cases.
Principales activités
Main tasks:
Additional tasks:
Avantages
Rémunération
Monthly gross salary : 2100 € during the first and second years. 2190 € the last year.