About The Workshop
Like the famous King Midas, popularly remembered in Greek mythology for his ability to turn everything he touched with his hand into gold,
we believe that the wealth of data generated by modern technologies, with widespread presence of computers, users and media connected by Internet, is a goldmine for tackling a variety of problems in the financial domain.
The MIDAS workshop is aimed at discussing challenges, potentialities, and applications of leveraging data-mining tasks to tackle problems in the financial domain.
The workshop provides a premier forum for sharing findings, knowledge, insights, experience and lessons learned from mining data generated in various application domains.
The intrinsic interdisciplinary nature of the workshop constitutes an invaluable opportunity to promote interaction between computer scientists, physicists, mathematicians, economists and financial analysts, thus paving the way for an exciting and stimulating environment involving researchers and practitioners from different areas.
Topics
- Trading models
- Discovering market trends
- Predictive analytics for financial services
- Network analytics in finance
- Planning investment strategies
- Portfolio management
- Understanding and managing financial risk
- Customer/investor profiling
- Identifying expert investors
- Financial modeling
- Anomaly detection in financial data
- Fraud detection
- Anti-money laundering
- Discovering patterns and correlations in financial data
- Text mining and NLP for financial applications
- Sentiment and opinion analysis for finance
- Financial network analysis
- Financial time series analysis
- Pitfalls identification
- Financial knowledge graphs
- Learning paradigms in the financial domain
- Explainable AI in financial services
- Fairness in financial data mining
- Quantum computing for finance
- Generative models for synthetic data
- Large language models in finance
- Agentic AI in finance
Important Dates
Paper Submission deadline: June 5, 2026
Acceptance notification: July 10, 2026
Camera-ready deadline: July 19, 2026
Workshop date: September 11 (morning), 2026
All deadlines are 11:59 PM AOE (Anywhere On Earth)
Submission
Submission Guidelines
Regular papers should refer to novel, unpublished work, and they can be either full or short. Full regular papers report on mature research works. Short regular papers include the following three categories:
- preliminary/work-in-progress research works
- demo papers
- survey papers
All the papers must be formatted according to the Springer LNCS style (https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines).
Regular papers may be up to 15 pages (full papers) or 8 pages (short papers). Extended abstracts may be up to 4 pages. All page limits are intended EXCLUDING REFERENCES, which may take as many additional pages as preferred.
Every paper should clearly indicate (as a subtitle, or any other clear form) the category it falls into, i.e., "full regular paper", "short regular paper", "extended abstract". As for short regular papers, we also require to provide the subtype, i.e., "short regular paper - preliminary", "short regular paper - demo", "short regular paper - survey". As for extended abstracts, we also require to specify whether it reports on some paper(s) already published and include the corresponding reference(s), i.e., "extended abstract - published work [REFERENCE(S)]", or if it is a position/vision paper, i.e., "extended abstract - position/vision".
Regular papers will be peer-reviewed, and selected on the basis of these reviews. Extended abstracts will not be peer-reviewed: their acceptance will be decided by the program chairs based on the relevance of the topics therein, and the adherence to the workshop scope.
For every accepted paper - both regular papers and extended abstracts - at least one of the authors must attend the workshop to present the work.
Contributions should be submitted in PDF format, electronically, using the workshop submission site at https://cmt3.research.microsoft.com/ECMLPKDDWT2026/. Specifically, please follow these steps:
1. Log-in to https://cmt3.research.microsoft.com/ECMLPKDDWT2026/
2. Select the 'Author' role from the drop-down menu in the top bar
3. Click on '+ Create new submission... ' button
4. Select 'MIDAS - The 11th Workshop on MIning DAta for financial applicationS'
All the submitted papers must be written in English and formatted according to the ECML-PKDD 2026 submission guidelines available at https://ecmlpkdd.org/2026/submissions-research-track/.
Proceedings
Regular papers will be included in the proceedings by default. As for extended abstracts, it will be given the authors the chance of either including or not their contribution in the proceedings.
Please, contact the organizers for questions related to the submission or participation.
Program
All times below are Central European Summer Time (CEST)
All papers are assigned a slot of 12 minutes for presentation + 3 minutes of Q/A.
Opening
Speakers: Workshop Organizers
Keynote
Certificates, Not Scores: Safety Guarantees for Machine Learning in Finance
Andrè Panisson, Principal Researcher, Intesa Sanpaolo AI Research
Abstract
Financial institutions increasingly run machine learning where an error is not a cost to be optimized but a compliance failure. In these settings the standard objective is unusable: minimizing expected loss demands a price for something regulation treats as a boundary, and a decision no auditor can reconstruct is undeployable at any level of accuracy. The object of interest is therefore not the classifier but the certificate: a bounded, auditable claim that a named population error rate stays below a named threshold. I take sanctions screening as the running case, where over 99% of alerts are false positives and every one is reviewed by hand. I situate the problem against what the literature offers: Neyman-Pearson classification, conformal and distribution-free risk control, selective prediction, learning to defer, and what each costs when the target rate is one in a million. I then present BP-NPC, a Bayesian partition-based method validated on twelve months of production data, which roughly doubles the volume safely auto-cleared at a fixed false-negative budget while keeping every decision traceable to explicit evidence. I close with the problem that success creates: a system that auto-clears stops observing what it clears.Short Paper
Monitoring Market Relationship Shifts with Time-Aligned Sliding-Graph Embeddings
Alex Romanova (Graph AI Studio)
Extended Abstract
TabPFN for Insurance Pricing: Do or Don't?
Bruno Deprez (KU Leuven), Wouter Verbeke (KU Leuven), Tim Verdonck (University of Antwerp)
Regular Paper
Predicting Reader-Interpreted Investment Stance Distributions from Financial Social Media Posts
Ren Hosokawa (Nara Institute of Science and Technology), Kentaro Ueda (Nara Institute of Science and Technology), Kota Tsubouchi (LY Corporation), Yuki Ogawa (Tokyo City University), Eiichi Umehara (Tokyo City University), Keiichi Yasumoto (Nara Institute of Science and Technology), Hirohiko Suwa (Nara Institute of Science and Technology)
Short Paper
Unsupervised Anomaly Detection Using Flow Matching on Tabular Data
Tejaswini Medi (University of Mannheim), Philip Konz (University of Mannheim), Margret Keuper (University of Mannheim)
Short Paper
Ceci n’est pas un client: Modelling Accounts via GINE for Remote Intrusion and Transaction-Takeover Estimation
Tarik Kalai (KU Leuven), Bruno Deprez (KU Leuven), Bertrand Lebichot (UC Louvain), Tim Verdonck (UAntwerp), Wouter Verbeke (KU Leuven)
Regular Paper
MyQuantitative: Risk-Aware Multi-Agent Portfolio Construction with Conformal Prediction
Angelo Thompson (Maastricht University), Filip Schlembach (Maastricht University), Misha Glazunov (Maastricht University), Evgueni Smirnov (Maastricht University)
Short Paper
Lightweight Post-hoc Refinement of Zero-shot Forecasts for Financial Time Series
Diogo Chaves (Universidade Federal de Minas Gerais), Pedro Robles (Universidade Federal de Minas Gerais), Turi Andrade (Universidade Federal de Minas Gerais), João Marcos Campos (Universidade Federal de Minas Gerais), Rafael Gomes (Universidade Federal de Minas Gerais), Gisele Pappa (Universidade Federal de Minas Gerais), Wagner Meira (Universidade Federal de Minas Gerais)
Closing Remarks
Speakers: Workshop Organizers
Registration
Please refer to the ECMLPKDD2026 Conference Website for registration details and instructions.
Workshop Location
University of Naples Federico II
Naples, Italy
The MIDAS2026 Workshop will be held in conjuction with the ECML-PKDD 2026 Conference, hosted at the University of Naples Federico II, in the Department of Electrical Engineering and Information Technology, located at Via Claudio 21, 80125 Naples, Italy.
Organizers
Ilaria Bordino
UniCredit, Rome (Italy)
Ivan Luciano Danesi
UniCredit, Milan (Italy)
Francesco Gullo
University of L'Aquila (Italy)
Domenico Mandaglio
University of Calabria (Italy)
Giovanni Ponti
ENEA (Italy)
Lorenzo Severini
UniCredit, Rome (Italy)
Program Committee
Luca Barbaglia, European Commission - Joint Research Centre
Ludovico Boratto, University of Cagliari
Cristian Bravo, University of Western Ontario
Federico Cini, Sapienza University of Rome
Andrea D'Angelo, University of Aarhus
Federico Gatta, Scuola Normale Superiore
Marco Gregnanin, IMT School for Advanced Studies Lucca
Daniel Benniah John, UC Berkeley
Wojciech Kurylek, University of Warsaw
Lucio La Cava, University of Calabria
Seungju Lee, Seoul National University
Stefano Francesco Monea, University of Calabria
Giulia Preti, CENTAI
Francesco Scala, CNR-ICAR
Reza Shahbazian, University of Palermo
Tik Yu Yim, University of Hong Kong
