Fault event detection through causal analysis of binary time series

Nokia is a global leader in the technologies that connect people and things. With state-of-the-art software, hardware and services for any type of network, Nokia is uniquely positioned to help communication service providers, governments, and large enterprises deliver on the promise of 5G, the Cloud and the Internet of Things.

Serving customers in over 100 countries, our research scientists and engineers continue to invent and accelerate new technologies that will increasingly transform the way people and things communicate and connect.

About Nokia Bell LabsOver its nearly 100-year history, Nokia Bell Labs has invented many of the foundational technologies that underpin information and communication networks and all digital devices and systems.

This research has resulted in nine Nobel Prizes, four Turing Awards, three Japan Prizes and a plethora of National Medals of Science and Engineering, as well three Emmys, two Grammys and an Oscar for technical innovations.

Nokia Bell Labs continues to conduct disruptive research focused on solving the challenges of the new digital era and innovating the technology that will define the next industrial revolution.

Internship : Fault event detection through causal analysis of binary time seriesIn general, RCA is a hard problem in complex systems, because it requires a deep knowledge of cause-effect dependencies among many features, physical and logical components the network nodes.

In the data driven approach, where most of this knowledge is assumed to be unavailable a priori, a major difficulty can emanate from the fact that many of the variables are hidden or unknown.

Furthermore, even in a fully observable system we are faced with the combinatorial explosion of cause-and-effect dependencies and the difficulty to collect enough information for distinguishing causal dependencies from spurious correlations.

The goal of this project is to explore techniques for causal analysis of binary (or discrete-valued, in a more general case) time series that represent local state change sequences of monitored network resources.

We aim at deriving a causal graph that explains the relationships between these time series, whereas the occurrence of an original (source or root) fault can be considered as an external intervention into the system on a particular network resource that can causally propagate to other resources.

More specifically, given as input a set of time series, our objectives is1.detection of fault occurrences helped by the inference of high-level causal graphs before and after these external interventions,2.

Study of different data generating models such as Noisy-OR or dynamic Bayesian networks, structural equation models adapted to time-series, or models derived from formal methods (event generating models based on connected automata on Petri nets)3.

Experimentations with the synthetic data as well as with real alarm logs collected in operator networks.The internship will be co-advised by Armen Aghasaryan (Nokia Bell Labs, Paris-Saclay) and Gregor Gössler (INRIA Grenoble).

Candidates must have a strong interest in formal methods (concurrency theory, Bayesian networks) as well as have good skills for data intensive simulation and analysis.

The trainee will benefit from a large degree of autonomy regarding the evaluation and interpretation of results as well as the tuning of the algorithm.

Outstanding work performed by the intern may lead to co-authorship in publications and can also be pursued in a PhD thesis.

Duration 6 months (full time)Qualifications : Last year Master-level student (final project). Solid technical skills and background in at least some of the followingareas are required :

  • Causal inference and modelling, Machine learning, Bayesian networks, Big Data analytics
  • Analytics platforms, Spark and Spark streaming, Hadoop / MapReduce
  • Python or Matlab Programming skillsLocation of the internship : Nozay 91620Site accessible from Massy RER station by public transport and Free buses departing from the Pont de Sèvres, Porte d'Orléans, Argenteuil, Chaville, Fontenay-le-fleury.
  • Imagine creating technology that has the potential to change the world. Working with us, you will have a positive impact on people’s lives and help to overcome some of the world’s most pressing challenges.

    We act inclusively and respect the uniqueness of people. At Nokia, employment decisions are made regardless of race, color, national or ethnic origin, religion, gender, sexual orientation, gender identity or expression, age, marital status, disability, protected veteran status or other characteristics protected by law.

    Nokia culture welcomes people as their true selves. Come create the technology to connect the world. Previous Job SearchesMy ProfileCreate and manage profiles for future opportunities.

    Go to ProfileMy SubmissionsTrack your opportunities.My SubmissionsStay in touch with us through our social media channels : Similar ListingsSW development traineeMN Mobile Networks Nozay, France, France?

    Applied R&DRequisition # : 21000004MOJobsHelpSite Map See all jobs Research & Development See new jobs Services See new jobs Market & Sales development See new jobs Sales See new jobs Finance See new jobs Corporate services See new jobs Nokia is an equal opportunity employer that is committed to diversity and inclusion.

    At Nokia, employment decisions are made regardless of race, color, national or ethnic origin, religion, gender, sexual orientation, gender identity or expression, age, marital status, disability, protected veteran status or other characteristics protected by law.

    Get in touch Linkedin Facebook Glassdoor Instagram Chat with our insiders Jobs Promote Jobs Analytics Help Site Map Cookies Privacy Terms 2019 Nokia Global Careers.

  • All rights reserved. var vMinLength '8';var vMaxLength '30';var vMinNonAlpha '0';var vUpperLowerRequired 'true';var vIsNotUserIdUserName 'true';
  • var vMinimumNumericCharacters '1';var vMinimumSpecialCharacters '1';var vMinLengthErrTxt 'Minimum number of characters in password is';
  • var vMaxLengthErrTxt 'Maximum number of characters in password is';var vMinNonAlphaErrTxt 'The password must contain at least';
  • var vNoChangePassword 'Please fill in your new password.';var vUpperLowerRequiredErrTxt 'The password must contain at least one uppercase letterand at least one lowercase letter.

    var vNoPasswordSpace 'No spaces.';var vIsNotUserIdUserNameErrTxt 'The new password may not be the same as your email address.

  • var vMinimumNumericCharactersErrTxt 'Minimum number of numeric characters in password is';var vMinimumSpecialCharactersErrTxt 'Minimum number of special characters in password is';
  • Special groups for clients who have aditional requirementsvar vSpecialGroup1 '';var vMinSpecialGroup1 '';var vMinSpecialGroup1ErrTxt 'You need 1 of the following characters .

  • var vSpecialGroup2 '';var vMinSpecialGroup2 '';var vMinSpecialGroup2ErrTxt 'You need 1 of the following characters .';var vSpecialGroup3 '';
  • var vMinSpecialGroup3 '';var vMinSpecialGroup3ErrTxt 'This is where you set the error given by special group #3'; / / Error Messages - optionalvar vErrHeaderDefault 'Please conform to the following criteria : ';
  • var vErrFooterDefault 'FieldError.PasswordErrorFooter';var vNoLoginPasswordTxt 'Please type in your existing password.

    var vJSDictCharacters 'characters.'; / / ' characters.'var vJSDictNumber 'non-alphabet character.'; / / ' number.'var vJSDictNumbers 'non-alphabet characters.

    numbers.' / / 'Please make sure that both of your new passwords match.'var vJSDictEquality 'Passwords must match.

    var vErrSubmitForm 'There was a problem submitting the form : '; / / Error messages for registration page checksvar vRequiredField 'Please fill in this field.

    var vAtLeast 'Length must be at least ';var vValidDateRequired 'Please fill in a valid date in this field using the format mm / dd / yyyy.

    var vEmailNotFormattedCorrectly 'Please provide a valid e-mail (e.g. person company.com). ';var vEmailDoesNotMatchConfirm 'Your FieldMap.

    Member.email and FieldMap.Member.emailConfirm do not match.';var vNoConsent1 'Please agree to the Privacy Policy.';var vNoConsent2 'Please agree to the Terms of Service.

    var vNoConsent 'Please agree to the Terms of Service.';var vAuth1or2Required 'To verify your account, please fill out FieldMap.

    Member.mb auth.auth info1 or FieldMap.Member.mb auth.auth info2 with at least 4 characters.';var vRegistrationSubmit 'FieldError.

  • RegistrationSubmit';function setGaRelationship() var ga relationship name 'unaffiliated';return ga relationship name; function setGaReferrer() var ga referrer relationship;
  • ga referrer relationship 'No referrer';return ga referrer relationship; var gaq gaq ;var ga relationship setGaRelationship();
  • var ga referrer setGaReferrer(); gaq.push( ' setAccount', 'UA-46796559-19' ); / / Site Specific Tracker gaq.push( ' setVar','Relationship : ' + ga relationship + ' ' + ga referrer );
  • gaq.push( ' setCustomVar',1,'Relationship',ga relationship,3 ); gaq.push( ' setCustomVar',2,'Referrer',ga referrer,3 ); gaq.

    push( ' trackPageview', window.location.pathname ); gaq.push( 'aggregate. setAccount', 'UA-19128950-16' ); / / Aggregate Tracker Alumni gaq.

  • push( 'aggregate. setDomainName', '.selectminds.com' ); gaq.push( 'aggregate. setVar','Relationship : ' + ga relationship + ' ' + ga referrer );
  • gaq.push( 'aggregate. setCustomVar',1,'Relationship',ga relationship,3 ); gaq.push( 'aggregate. setCustomVar',2,'Referrer',ga referrer,3 );
  • gaq.push( 'aggregate. trackPageview', window.location.pathname );var ga alert '';if(ga alert 'yes') alert('Page View : ' + window.

  • location.pathname + ' Relationship : ' + ga relationship + ' Referrer : ' + ga referrer); (function() var ga document.createElement('script');
  • ga.type 'text / javascript'; ga.async true; ga.src ('https : ' document.location.protocol ? 'https : / / ssl' : 'http : / / www') + '.

  • google-analytics.com / ga.js'; var s document.getElementsByTagName('script') 0 ; s.parentNode.insertBefore(ga, s); )();function asTrackPageview(url name) var ga relationship name setGaRelationship();
  • var ga referrer relationship setGaReferrer(); gaq.push( ' setCustomVar',1,'Relationship',ga relationship name,3 ); gaq.push( ' setCustomVar',2,'Referrer',ga referrer relationship,3 );
  • gaq.push( ' setVar','Relationship : ' + ga relationship name + ' ' + ga referrer relationship ); gaq.push( 'aggregate. setCustomVar',1,'Relationship',ga relationship name,3 );
  • gaq.push( 'aggregate. setCustomVar',2,'Referrer',ga referrer relationship,3 ); gaq.push( 'aggregate. setVar','Relationship : ' + ga relationship name + ' ' + ga referrer relationship );
  • if(ga alert 'yes') alert('Page View : ' +url name + ' Relationship : ' + ga relationship name + ' Referrer : ' + ga referrer relationship);
  • gaq.push( ' trackPageview', url name ); gaq.push( 'aggregate. trackPageview', url name ); / / alert('Tracking Page : ' + url name);
  • function asTrackEvent(category, action, opt label) var ga relationship name setGaRelationship();var ga referrer relationship setGaReferrer();
  • if(ga alert 'yes') alert('Event : ' +category +'.'+ action +'.'+ opt label + ' Relationship : ' + ga relationship name + ' Referrer : ' + ga referrer relationship);
  • gaq.push( ' trackEvent', category, action, opt label ); gaq.push( 'aggregate. trackEvent', category, action, opt label );
  • alert('Tracking Event : ' + category + ' ' + action + ' ' + opt label); 00

    Depuis le 8 Mai 2021

    Fermé le 16 Août 2021

    Appliquer

    Fault event detection through causal analysis of binary time series

    Stage/Apprentissage: Ingénieur RD SW 5G

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series

    Fault event detection through causal analysis of binary time series