October 2025 - SIPTA News - lists.sipta.org

3 years Postdoc position at TCD (Dublin) on Security & Privacy of unused parameters of ML models.

by Alessio Benavoli

Dear colleagues and students, I am recruiting a Postdoctoral researcher with a background in game theoretic modelling, learning theory and both the theoretical and practical aspects of deep learning. The successful applicant will collaborate within a skilled and motivated team of scientists (from Trinity College Dublin, Queen's University Belfast and University California Riverside) dedicated to understand the security and privacy implications of the unused parameters present in machine learning models. The proposed research explores the problem both from the theoretical and empirical perspectives. *Description*: Machine Learning applications increasingly rely on Deep Neural Networks (DeepNNs) to achieve state-of-the-art performance across a wide range of real-world tasks, including Large Language Models. DeepNNs are typically highly over-parameterised, but this characteristic has been shown to facilitate both optimisation and generalisation. Nevertheless, such over-parameterisation also introduces many parameters that are not essential to the final learned model and whose specific values have little impact on its behaviour (don’t-care states). In both software and hardware systems, undefined behaviour and don’t-care states have been shown to be potential sources of vulnerabilities. The postdoc will conduct research that contributes to the following streams of work: 1) Investigate the possible strategies that colluding parties could use to maliciously exploit the unused parameters (learning capacity) of a machine learning model using techniques from information and learning theory; 2) Model the attacks within a game-theoretic framework under complete or incomplete knowledge (Bayesian games), which will enable to find optimal strategies for the players (Attacker, Nature and Defender); 3) Validate these theoretical models and strategies on machine learning models and real applications. Applications from candidates with expertise in *imprecise** Statistics or imprecise Machine Learning* are welcome. - For informal *inquiries* please contact *alessio.benavoli(a)tcd.ie <alessio.benavoli(a)tcd.ie>* - *Salary* is up to ~€71,000 per annum per 3 years. - *Start*: January 2026. - For additional details (including qualifications, knowledge/experience and how to apply)* see attached document.* Best regards, Alessio Benavoli

Oct. 29, 2025

CFP IPMU 2026- Special Session on Theoretical and applied aspects of Imprecise Probabilities

by IGNACIO MONTES GUTIERREZ

Dear colleagues, Please find attached a CFP for a special session on IP at the IPMU 2026 Conference, which will be held in Rome in June 2026. Best regards Ignacio Montes ****************************************************** [Apologise for multiple copies] Call for papers Special Session on "THEORETICAL AND APPLIED ASPECTS OF IMPRECISE PROBABILITIES" to be held at the 21st International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems (IPMU 2026) https://www.sbai.uniroma1.it/conferenze/ipmu2026/index.php Rome (Italy), June 15-19, 2026 Session topic and goal This session is devoted to Imprecise Probability Theory, which involves all the mathematical models that can be used as more flexible tools than usual Probability Theory when the available information is scarce, vague or incomplete. It includes lower previsions, n-monotone capacities, belief functions, p-boxes, possibility measures, or non-additive measures, among others. This special session aims to include papers related to Imprecise Probabilities that either present a significant advance in the foundations or show potential applications in real problems. In addition, papers where the connection between imprecise probability theories and other fields such as fuzzy sets or game theory is emphasised are also welcome. Theoretical topics of interest include (but are not limited to): * (coherent) uncertainty models * non-additive set-functions * qualitative reasoning about uncertainty * limit laws for imprecise probabilities * philosophical foundations * elicitation and inference * robust statistics * decision making Applications include (but are not limited to): * data mining * classification and machine learning * risk and reliability analysis * finance * life sciences * system control and design * algorithmic issues Dates Paper submission: January 12, 2026 Notification of Acceptance: February 15, 2026 Conference: June 15-19, 2026 Organisers We would be grateful if you could let us know whether you plan to submit a paper to this session. Any questions or remarks can be addressed to: David Nieto-Barba (nietodavid(a)uniovi.es) University of Oviedo (Spain) Ignacio Montes (imontes(a)uniovi.es) University of Oviedo (Spain)

Oct. 18, 2025

Reminder: SIPTA seminar at 3pm Paris time

by sdesterc

Dear Siptians, This a small reminder that our seminar is bound to start in a bit more than one hour. Milan Studeny will talk of “Combinatorial description of facets of the cone of exact games (or coherent lower probabilities)”. Zoom link is https://utc-fr.zoom.us/j/85055146633 Best regards Sébastien -- ==================================== Sebastien Destercke, Ph. D. CNRS researcher in computer science. Université de Technologie de Compiegne U.M.R. C.N.R.S. 7253 Heudiasyc Avenue de Landshut F-60205 Compiegne Cedex FRANCE Tel: +33 (0)3 44 23 79 85 Fax: +33 (0)3 44 23 44 77 ====================================

Oct. 17, 2025

CNRS positions - reminder

by sdesterc

Dear colleagues, As each year, the CNRS (centre national de recherche scientifique) will hire a number of people that will work in one of its associated laboratories (nation-wide in France), requiring candidates to provide a research project as well as a list of possible laboratories (associated to CNRS) in France where they could perform the proposed research. * Why should you apply? - CNRS positions are full permanent research positions that are, to my knowledge, one of the most flexible in terms of how you do your research. - CNRS offers quite a lot of opportunities (e.g., international labs) in case you want to make long foreign stays or develop your research into a product at some point. - As a CNRS researcher, you are mostly free to join any joint CNRS lab. This means that moving within France is very easy. - The institute is committed (and I can tell that by experience) to perform qualitative rather than quantitative assessments. So no obligation to publish or perish to prosper, but you can still play the game if you want. * Why should you not apply? - The salary is a priority for you: while it is quite decent at the start and quickly increases (in a rather predictable manner, as for all French state employees), it is probably below the level of tenured people in other European countries, notably due to the fact that France has historically chosen to have a lot of permanent public researchers/lecturers/professors (so monetary means are not concentrated nor subject to competition between universities). There is a number of ways to increase it (work part-time in company, give hours of teaching, providing some expertise, having salary supplements within projects), but most take some time away from research. - Learning French is not an option for you: while CNRS hires on a very regular basis non-french speaking researchers, daily life in France is definitely much more difficult if you do not have a working French. This can typically be acquired within some years (or maybe less) after arrival. * How should you apply? - Calendar: application files are typically to be submitted around January, with interviews in May. In order to prepare the candidacy, it is advisable to take contact with prospective teams/laboratories in October, November at the latest (preparing the application files take quite some time, and should be done in coordination at least with your preferred team) - Application file: It consists of a CV, a research statement (of about 10 pages in average), and of a research project that explains your scientific project for the next 5/10 years, understandable for specialist of the section you apply to (in computer science, the committee includes people from artificial intelligence, quantum computing, logic, mathematical programming, …). Within the project must be given three CNRS laboratories and teams (to be selected by the candidate among all the possible ones) in which you would like to perform this project. - Contacting laboratories/teams: if you have a laboratory/team you would like to join in case you are selected by CNRS, you should contact it as soon as possible. If you are interested or if you think of someone that could be interested, I will be happy to answer further questions, for instance to recommend labs where you could apply depending on your research project. Best regards Sébastien

Oct. 16, 2025

FW: IEEE WCCI 2026 Maastricht | First Announcement

by Kreinovich, Vladik

IEEE WCCI 2026 - 21-26 June 2026, Maastricht, the Netherlands https://attend.ieee.org/wcci-2026/ Welcome We are pleased to announce that the 2026 IEEE World Congress on Computational Intelligence (WCCI 2026) will be held in Maastricht, the Netherlands, from 21 to 26 June 2026. This congress brings together three flagship conferences, facilitating rich knowledge exchange in neural networks, fuzzy systems and evolutionary computation: * 2026 International Joint Conference on Neural Networks (IJCNN) * 2026 IEEE International Conference on Fuzzy Systems (FUZZ) * 2026 IEEE Congress on Evolutionary Computation (CEC) Call for Submissions We look forward to receiving many outstanding research contributions for the 2026 IEEE World Congress on Computational Intelligence. Help shape the event by submitting a high-quality research paper, organizing a cutting-edge competition or challenge, proposing a focused special session, leading an interactive hands-on tutorial, hosting a collaborative thematic workshop, or joining a thought-provoking discussion panel as a participant or moderator. Call for Papers Paper submissions are invited (from 1 December 2025) via the WCCI 2026 website. All papers will be peer-reviewed and evaluated based on originality, significance, quality, and clarity. https://attend.ieee.org/wcci-2026/call-for-papers/ Special Sessions Special Session proposals should clearly outline the research domain of the session, highlighting its significance, impact, and relevance to the conference and the wider community. https://attend.ieee.org/wcci-2026/call-for-special-sessions/ Call for Tutorials We invite you to propose high-quality Tutorials and Workshops that align with the theme of WCCI. These will be pivotal to enhancing the learning experience of attendees. https://attend.ieee.org/wcci-2026/1222-2/ Important dates 1 November: Special sessions, competitions, tutorials, panels proposals 21 November: Special sessions, competitions, tutorials, panels notification 1 December 2025: Submission system opens 31 January: Paper submission deadline (23:59, anywhere on Earth, i.e. UTC-12) - no extension will be given! 15 March: Paper acceptance notification 15 April: Camera-ready papers 21-26 June: Congress Days Become a partner of WCCI 2026! The success of the World Congress on Computational Intelligence relies on the generous support of our esteemed sponsors. We invite you to join us as a sponsor and become an integral part of 2026 WCCI. Please find detailed information about our sponsorship opportunities at https://attend.ieee.org/wcci-2026/sponsors/ Organized by IEEE WCCI 2026 is legally organized by Stichting ConferEntis on behalf of IEEE, IEEE Computational Intelligence Society, International Neural Network Society, Maastricht University, and the University of Nottingham. Klinkhamer Group B.V. has been contracted by Stichting ConferEntis to manage the conference organization and secretariat. Klinkhamer Group B.V. has its registered office at: Duboisdomein 5D01, 6229 GT Maastricht, The Netherlands. E: wcci2026(a)klinkhamergroup.com<mailto:wcci2026@klinkhamergroup.com> T: +31 (0)43 362 70 08

Oct. 15, 2025

Call for Special Sessions IPMU2026. Rome, Italy, June 15th-19th, 2026. Extended Deadline

by Kreinovich, Vladik

*********************************************************** 21st International Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems (IPMU 2026) June 15th-19th, 2026, Rome, Italy https://www.sbai.uniroma1.it/conferenze/ipmu2026/ ************************************************************ The IPMU conference is organized every two years with the aim of bringing together scientists working on information processing and the management of the many aspects of uncertainty. It also serves as an ideal forum for the exchange of ideas between theoreticians and practitioners in these and related areas. The 2026 edition of IPMU will take place at the Faculty of Economics of Sapienza University of Rome, Italy, in the nearby of the University city of Rome and the main train stations (Termini and Tiburtina). The 2026 edition of IPMU celebrates the 40th anniversary from the first edition of the conference that took place in Paris, France. The choice of Rome for this special occasion is not accidental since Rome and Paris are exclusively and reciprocally twinned, starting from April 1956. Sapienza University of Rome is one of the oldest universities of the world, being it founded in 1303 by Pope Boniface VIII, and is the largest university in Europe. Due to its strategic position, the Faculty of Economics allows to easily reach the plethora of monuments, historic hot-spots and wonderful sightseeing places that awarded to Rome the name of the Eternal city. Get ready for discussing cutting-hedge research on methods for the management of uncertainty and aggregation in one of the most beautiful cities of the world. On behalf of the Organizing Committee of the 21st IPMU Conference, we invite researchers to submit original research contributions (theoretical, methodological, applications) and/or to organize a special session on a specific topic within the scope of the conference which includes (but is not limited to) the following topics: Theory, Methods and Tools: Measures of Information and Uncertainty Bayesian and Probabilistic Methods Evidence, Possibility and Utility Theories Imprecise Probabilities Fuzzy, Rough, Intuitionistic, etc, Sets and operators Fuzzy Logic and other non-classical Logics Multiple Criteria Decision Methods Aggregation Methods Knowledge Acquisition, Representation and Reasoning Graphical Models Machine Learning Evolutionary Computation Neural Networks Data Analysis and Data Science. Application Fields: Big Data Smart Cities Image Processing Intelligent Systems and Information Processing, Logistics, Transportation and Routing Information Retrieval and Fusion Agents E-Health, Medicine and Bioinformatics Finance Fuzzy Optimization Please note that special sessions usually consist of about 4-6 presentations. You should send your proposal to ipmu2026(a)sbai.uniroma1.it<mailto:ipmu2026@sbai.uniroma1.it> Please, include in the message a description, list of topics and organizers. IMPORTANT DATES Special sessions proposals extended deadline: October 22nd, 2025 Notification of Special Sessions acceptance: October 31st, 2025 General Chair Barbara Vantaggi (Italy) Program Chairs Giulianella Coletti (Italy) Thierry Denoeux (France) Anne Laurent (France) Davide Petturiti (Italy) Publication Chair Enrique Miranda (Spain) Special Session Chair: Jesús Medina (Spain) Social Media Chairs Andrea Cinfrignini (Italy) Silvia Lorenzini (Italy) Executive Directors Bernadette Bouchon-Meunier (France) Ronald R. Yager (USA) [Image removed by sender. UCA] Jesús Medina Moreno Catedrático de Universidad / Full Professor Mathematics for Computational Intelligence Systems (M·CIS) Departamento de Matemáticas Universidad de Cádiz Facultad de Ciencias, Campus del Río San Pedro 11510 Puerto Real Cádiz - SPAIN Tel +34 956 01 27 12

Oct. 15, 2025

mark your calendars

by Kreinovich, Vladik

8th International Conference on Intelligent and Fuzzy Systems (INFUS 2026) July 7-9, 2026 Ankara, Türkiye infus.itu.edu.tr Conference Theme: Artificial Intelligence for a Sustainable Future: Driving Innovation in the Circular Economy Conference Venue: Cankaya University, Ankara, the capital city of Türkiye Conference Topics: Approximate reasoning, Artificial Neural Networks, Autonomous Systems, Classification, Clustering, Cognitive modelling, Computational Finance and Economics, Computing with Words, Data aggregation and fusion, Database management and querying, Decision-Making, e-Business, e-Learning, Evolutionary/Genetic Algorithms, Expert Systems, Feature Extraction, Forecasting and time series modelling, Fuzzy control, Fuzzy Systems, Geographic Information Systems, Hybrid systems involving software agents and human actors, Image processing and computer vision, Imprecise probabilities and fuzzy methods in statistics, Information retrieval, Intelligent data analysis and data-mining, Intelligent Systems, Knowledge representation and knowledge engineering, Linguistic modelling, Machine Learning, Natural language processing, generation and understanding, Nature-Inspired Optimization Algorithms, Neuro-fuzzy systems, Optimization, Possibility theory and applications, Rough sets theory, Semantic web, Social Networks Analysis and Computing, Soft Computing Applications, Stochastic and fuzzy optimization, System Control and Optimization, Theoretical foundations of fuzzy logic and fuzzy set theory, Theory and applications of decision-making, Uncertainty modelling. Social Events: Visit to Gordion Ancient city, Visit to Antic Roman Baths, and Beypazarı Your registration involves * Presentation of one accepted paper at the INFUS 2026 Conference, * Publication of one accepted paper (within 8 pages) in the Springer proceedings (guaranteed only for registration received before April 05, 2026), * Access to remote / onsite tutorial sessions at the conference, * Access to all remote INFUS 2026 conference scientific sessions, * Travels to Gordion Antic City, Roman Baths, and Beypazari * Conference bag and gifts, * Gala dinner, welcome cocktail, three days coffee breaks, * Three days lunches * And more than all, a warm hospitality that will make you feel at home. Important Dates: * Final date to submit abstracts: January 25, 2026 * Notification of abstracts acceptance: February 08, 2026 * Final date to submit full papers: February 15, 2026 * Notification of full papers acceptance: March 1, 2026 * Early registration: March 08, 2026 * Camera-ready paper submission: March 22, 2026 * Source file submission : March 29,2026 * Deadline of author late registration for paper publication in proceedings: April 05, 2026 * Conference: July 07-09, 2026 Conference Chair: Prof. Cengiz Kahraman kahramanc(a)itu.edu.tr<mailto:kahramanc@itu.edu.tr> Conference Co-Chairs: Prof. Basar Oztaysi, Prof. Selcuk Cebi, Prof. Sezi Cevik Onar, Prof. Irem Ucal Sari, Prof. Abdullah Cagri Tolga, Assoc. Prof. Irem Otay

Oct. 11, 2025

SIPTA virtual seminar (game theory and lower probabilities): 17th October, 3pm Paris time

by sdesterc

Dear colleagues, We are delighted to announce our upcoming SIPTA online seminar on imprecise probabilities (IP). These monthly events are open to anyone interested in IP, and will be followed by a Q&A and open discussion. They also provide an occasion for the community to meet, keep in touch and exchange between in-person events. The next seminar will take place on the 17th of October (Thursday). The zoom link for this seminar is https://utc-fr.zoom.us/j/85055146633 For this new seminar, we are very happy to have Milan Studeny as our speaker. Milan Studeny is a senior research er at the Institute of Information Theory and Automation, within the Academy of Sciences of the Czech Republic. Milan Studeny is internationally recognised, notably for his contributions to graphical models and conditional independence, as well as to game theory, on which this SIPTA seminar will focus. On the 17th of October, at 15:00 CEST:paris time (up to 17:00 CEST, with a talk duration of 45min/1h), he will talk about "Combinatorial description of facets of the cone of exact games (or coherent lower probabilities)”. Curious? Then check out the abstract on the webpage of the SIPTA seminars: sipta.org/events/sipta-seminars. The zoom link for attending the seminar can be found on that same page. So please mark your calendars on the 17th of October, and join us for the occasion. And for those who missed the previous seminar and want to catch up, or simply want to see it again and again, it is online at https://www.youtube.com/watch?v=hLdCKytLbhY See you at the seminar! Sébastien, Enrique and Jasper

Oct. 7, 2025

Numbers of the Future Conference: **Submit the abstract of your presentation**

by Kreinovich, Vladik

Numbers of the future Conference When: 3-5 December 2025 Where: University of Liverpool, UK (also online for those who cannot travel) Registration: £100 (£50 online) Abstract submission: https://forms.gle/c1p1Jcikve1cjQET8 Conference website: https://sites.google.com/view/numbersofthefuture Taglines: https://sites.google.com/view/numbersofthefuture/taglines Overview Next-generation science and engineering will require data and calculations to be reliable, trackable, reusable, comprehensively described and linked, while also maintaining security and protecting private information. Quantitative assessments important in science, engineering, industry, commerce, finance, and forensics increasingly make use of critical data and calculations, sensitive or confidential records, incomplete or bad data, intrusively collected personal information, and tracked data and calculations. This use requires collating and propagating ancillary information that justifies conclusions, tracks evidence, and permits reanalysis. This conference will consider several essential questions: § What details are needed when collecting and sharing observational data and calculations? § How can we be sure that empirical effort produces meaningful, correct, and logically complete measurements? § What changes or augmentations are needed in empirical work to ensure that estimates and calculations are trackable and checkable? § What structures and processes can make scientific statements trustworthy, beyond relying on the integrity of scientists as individuals? § How can these be automated to be natural and least burdensome? § How can we radically reduce the need to 'clean' data sets? § What provisions can guarantee that the information is secure and uncorrupted but also accessible and reusable. § How can we avoid wasting or losing observations? Topics bad data, reusable data, symbolic data, trustworthiness, literate computing, data interoperability, responsible analysis, empiricist responsibility, linguistics of evidentiality, anonymisation techniques, numbers in crisis situations, findable and recoverable data, protecting personal information, self-documentation for data sets, evidence and justification tracing, uncertainty from expert elicitations, responsible simulation and modelling, ethical dimensions of numbers and data, security and confidentiality in data sharing, decision making under justification tracking, provenance and calculation stream histories, enriched and structured next-gen commenting, dimension and units checking and propagation, dangers and limitations in certification by analysis, extended research credits for non-author contributions, what to do about numbers already recorded in data sets, automating uncertainty propagation and sensitivity analysis, and conventions for comprehensive data description (NUSAP, FAIR, etc.). Themes * Reproducible and open science * Humane algorithms * Data sharing * Verification Who should attend The conference will bring together engineers with computer scientists, statisticians and data analysts, linguists, legal scholars, and policy analysts from across industrial, academic, and governmental institutions who must handle mission-critical data and calculations, sensitive or personal records, incomplete or bad data, or tracked data and calculations important in science, engineering, industry, commerce, finance, and forensics. Using such data requires the collation and propagation of ancillary information that justifies the quantitative calculations and assessments in which they are used. This multidisciplinary workshop is intended to foster cross-fertilization and creativity. Conference agenda The conference will begin Wednesday 3rd December at 2pm. A detailed agenda will be updated soon on the conference website. Wednesday (afternoon) Thursday Friday Registration Registration will open soon. In the meantime you can submit the abstract of your presentation at https://forms.gle/c1p1Jcikve1cjQET8. £100 fee for in-person registration will cover: § Attendance § Lunch and refreshment breaks, and § Conference dinner. On-line attendance has a £50 fee. Additional tickets (£50 for accompanying persons includes the conference dinner, and social events) may be purchased at registration. Important dates 31st October: deadline for abstract submission 1st November: registration opens 14th November: notification of acceptance of presentation 3rd December: conference January: special issue open Keynote speakers Arnald Puy<https://www.arnaldpuy.com/> (University of Birmingham) Ivan Oransky<https://retractionwatch.com/meet-the-retraction-watch-staff/about/> (Retraction Watch) TBC Journal special issue Selected abstracts/presentations from the conference will be invited for submission to a special issue of the International Journal of Approximate Reasoning. Conference venue The conference will be held in person at the University of Liverpool in the <https://www.liverpool.ac.uk/central-teaching-hub/> Central Teaching Hub<https://www.liverpool.ac.uk/central-teaching-hub/> (What3Words ///basin.causes.select<https://what3words.com/basin.causes.select>). The presentations will also be live streamed for registrants via Zoom. Organizers The conference is hosted by the Institute for Risk and Uncertainty at the University of Liverpool in the United Kingdom. Co-chairs Marco de Angelis<https://www.strath.ac.uk/staff/de_angelismarcodr/>, University of Strathclyde, United Kingdom Scott Ferson<https://sites.google.com/site/scottfersonsite/>, University of Liverpool, United Kingdom Kari Sentz<https://laro.lanl.gov/esploro/profile/kari_sentz/output/all?institution=01L…>, Los Alamos National Laboratory, United States Conference Secretariat Andrea Jones +44 0151 7944837 a.m.jones(a)liverpool.ac.uk<mailto:a.m.jones@liverpool.ac.uk> Steering Committee Ekaterina Auer, Hochschule Wismar, Germany Anas Batou, University of Liverpool, United Kingdom Michael Beer, Leibniz Universität Hannover, Germany Conal Brown, TÜV Rheinland Industrial Services, United Kingdom Mark Burgman, University of Hawai‘i at Mānoa, United States Leslie Yu Chen, University of Liverpool, United Kingdom Leonardo Christo, Agência Nacional do Petróleo, Gás Natural e Biocombustíveis, Brasil Silvio Funtowicz, Universitetet i Bergen, Norway Ander Gray, Université de Technologie de Compiègne, France Ioanna Ioannou, University College London, United Kingdom Vladik Kreinovich, University of Texas at El Paso, United States Adolphus Lye, National University of Singapore, Singapore Enrique Miralles-Dolz, Princeton Plasma Physics Laboratory, United States Jeremy NIcholls, Social Value International, United Kingdom William L. Oberkampf, United States Jason O'Rawe, Etsy, United States Edoardo Patelli, University of Strathclyde, Scotland Jerome Ravetz, University of Oxford, United Kingdom Sebastian Timme, University of Liverpool, United Kingdom Local Organizing Committee Ioanna Ioannou Nicholas Gray Leslie Yu Chen Conal Brown To know more about hotels, transportation to reach Liverpool and related ideas see the conference site: https://sites.google.com/view/numbersofthefuture.

Oct. 6, 2025