Bayesian and Dempster-Shafer models for combining multiple sources of evidence in a fraud detection system
By Fabrice Daniel, Artificial Intelligence Department of Lusis, Paris, France
Combining evidence from different sources can be achieved with Bayesian or Dempster-Shafer methods. The ﬁrst re-quires an estimate of the priors and likelihoods while the second only needs an estimate of the posterior probabilities and enables reasoning with uncertain information due to imprecision of the sources and with the degree of conﬂict between them. This paper describes the two methods and how they can be applied to the estimation of a global score in the context of fraud detection.
Fraud detection mainly relies on expert driven methods that implement a set of rules and data driven approaches implementing machine learning (ML) models. Both provide an estimate (or a score) for a new transaction to be fraudulent.
While each ML model naturally returns a fraud probability, the experts can also attach a probability to each rule. They can also be automatically calculated from the labeled his-tory. Combining them together produces a global score that can be used in a near real time system to rank a set of trans-actions having the highest probability to be fraudulent. By obtaining this ranking, investigators can concentrate their efforts on the suspect transactions with the highest probability of being true frauds.
The most common approaches for combining scores are summing individual scores or returning the highest score among the trigged rules. This is not entirely satisfactory given that summing scores is equivalent to averaging the probabilities returned by each predictor (rule or model). It also does not take into account the uncertainty of each predictor and the degree of conﬂict between them.
For the Lusis fraud system, we work on implementing more appropriate approaches.
This paper proposes two ways for addressing this problem. The ﬁrst is to use Bayesian methods ; the second is to combine the scores by using Dempster-Shafer theory .
As the end of the year draws near we would like to take the opportunity to thank all of our clients, partners and teammates for an extraordinary year, amidst the challenges of the global coronavirus pandemic which has affected families, businesses and communities. Our thoughts go out to anyone who has been impacted by the virus especially those who are sick, we extend our heartfelt wishes for a full recovery.
We have always said “we hold our clients central to our thinking and our actions” and that statement holds true this year more than ever. The success of Lusis Payments is based on the relationships we have built over the years, and we really wouldn’t be where we are without you. Over this past year, as more and more North American financial institutions have called upon Lusis Payments for our technology and services we added our first North American support center in Toronto Canada and have made NA our primary market. Until this year, Lusis had only sales and consulting teams in North America. The addition of this facility and the team assembled will dramatically accelerate new and current project deployments of TANGO.
The need for a cutting-edge, real-time fraud solution has never been so critical as it is now. The world has seen an unprecedented switch from CP (cardholder present), CNP (cardholder not present) to online transactions, which has massively changed the patterns of fraudulent activity for OLTP (online transaction processing). Our clients that process online payments needed the agility to rapidly counteract this industry swing. This year Lusis introduced TANGO AI Fraud (Tango AIF™) to provide a leading-edge solution for fraud monitoring, detection and action — technology that provides the latest methods for processing secure transactions, irrespective of source, nature or market segment. The development has followed years of dedication by our Artificial Intelligence department, created in 2017. As part of our continued dedication to AI, earlier this year Lusis created a research chair in partnership with CentraleSupélec a prestigious French graduate engineering school of Paris-Saclay University. Through this partnership, Lusis and CentraleSupélec are strengthening their collaboration in the field of artificial intelligence applied to the banking sector.
This summer Lusis also announced further expansion as we increased our presence in Mexico and select countries of Latin America. We were excited to introduce Mercedes Fabila as the newest addition to our sales executive team. Fabila joined the company as the Director of Sales, Mexico and Latin America.
Our most exciting news for the year was the new release of TANGO Version 8. TANGO V8 is a major upgrade that provides full microservice integration and complete Cloud capacities. The new features bring expanded capacities and even greater flexibility for delivery. Please contact us directly for a one-on-one presentation of TANGO’s full capabilities and new capacities.
Recently, we were able to interact with several clients, partners and friends during the Virtual NonStop Technical Bootcamp. Through zooms or instant messages, we were happy to answer questions and be able to say hello to colleagues in the industry. Our presentation “Today’s Payments Market: Creating Opportunities from Challenges” is still available for viewing through the Whova app.
As the new year approaches, we know there will be many new challenges ahead. We welcome all that 2021 has in store for us with enthusiasm and anticipation for all the opportunities around the world.
From the whole team at Lusis Payments we wish you all Happy Holidays!
CREATION OF THE RESEARCH CHAIR "ARTIFICIAL INTELLIGENCE APPLIED TO THE DETECTION OF PAYMENT FRAUD AND TRADING"
Romain Soubeyran, Managing Director of CentraleSupélec and Philippe Préval, Managing Director of Lusis, launch the Chair "Artificial intelligence applied to the detection of payment fraud and trading". Through this partnership, Lusis and CentraleSupélec are strengthening their collaboration in the field of artificial intelligence applied to the banking sector.
Lusis is the publisher of TANGO, a high-performance payments transaction platform for the finance industry. With the TANGO platform, Lusis provides a complete retail payment system that includes fraud detection, as well as extremely rich and complex “front to back” trading platforms.
In order to give a formal framework to this permanent R&D approach, in June 2017 Lusis created an Artificial Intelligence department responsible for working on Machine Learning / Deep Learning approaches mainly applicable to fraud detection on payments and trading strategies in the financial markets.
For its part, CentraleSupélec gives very high priority to artificial intelligence by dedicating focus to a of 3rd year to AI. The laboratories of CentraleSupélec are also present in this field by participating in numerous projects. This chair is the first signed by CentraleSupélec with a French high tech SME.
Strong involvement of students and researchers
Thanks to the long-standing collaboration between Lusis and CentraleSupélec, the student-engineers of the School have the opportunity to carry out, within the framework of "industrial study contracts" (« contrats d’étude industrielle » CEI), projects that focus on solving complex problems using statistical or machine learning approaches.
Involving students allows Lusis to generate and capture new ideas created in the laboratory that are usable in industrial approaches, and to access a pool of skills (researchers, students and collaboration network) conducive to the development of the various projects carried out.
Scientific objectives and research themes
The Chair supported by Lusis and CentraleSupélec provides a structured working framework allowing the two partners to carry out research work of specific interest by relying on a dedicated research team. It is housed in the Computer Science Research Laboratory (LRI), a joint research unit at Paris-Saclay University, CNRS and CentraleSupélec.
The themes studied include the detection of fraud on payments and trading strategies.
As far as fraud detection is concerned, the essential points to take into account are both the appearance of new fraud strategies and the essentially online nature of the process, as well as the ability to provide an explanation for the refusal or the acceptance of a transaction.
With regard to trading strategies, the work undertaken highlights the share of random time series in financial market data. This is the major difficulty of research in this area.
"Artificial intelligence is a game changer in the world of IT and business in general. It will completely redraw the landscape of whole sections of the industry. We want Lusis to be part of this revolution, particularly in our preferred areas of trading, payments systems, on-line banking, and customer-centric programs. For us, being an AI player means being a stakeholder in AI research. This is why, we are both happy and proud to be able to give a new dimension to our long-standing partnership with CentraleSupélec. Our teams like those set up by CentraleSupélec are enthusiastic about exploring these new universes,” declares Philippe Préval, CEO of LUSIS.
“Artificial intelligence is a young science whose recent successes and advances open up enormous potential for innovation in all fields and sectors of activity, which is at the heart of CentraleSupélec's strategy. We train, through dedicated courses, the engineers and researchers who will build tomorrow's intelligent systems at the service of society. With a wide disciplinary variety, our research activities cover a wide spectrum of AI," says Romain Soubeyran, Managing Director of CentraleSupélec who adds"The artificial intelligence chair applied to the detection of payment fraud and trading with LUSIS allows us to reinforce this strategy with a very ambitious scientific program.”
CentraleSupélec is a public scientific, cultural and professional establishment, born in January 2015 of the merger of École Centrale Paris and Supélec. Today, CentraleSupélec consists of 3 campuses in France (Paris-Saclay, Metz and Rennes). It has 4,300 students, including 3200 engineering students, and brings together 16 laboratories or research teams. strongly internationalized (30% of its students and almost a quarter of its international teaching staff), the school has established over 170 partnerships with the best institutions in the world. Leading school in higher education and research, CentraleSupélec is a benchmark center in the field of engineering and systems sciences, ranked among the best institutions world. It is a founding member of Paris-Saclay University and chairs the Groupe des Ecoles Centrale (Lyon, Lille, Nantes and Marseille), which operates international establishments (Beijing (China), Hyderabad (India), Casablanca (Morocco)).
Lusis is a technology supplier allowing to implement solutions of:
• Payment HUB
• Advanced messaging
• Fight against fraud
Lusis payments is a major player in the payments world. We focus on R&D, innovation, performance, time to market, disruption. We
take advantage of SCRUM both for our developments and for the projects we carry out with our customers. The architecture of our micro-service (Tango) solution and benefits from a constant investment of R&D on new payments, AI, distributed ledgers, performance.
We have clients all over the world, operating directly or through partnerships from:
• Paris: HQ, R&D
• London: EMEAA Sales
• San Francisco: America’s sales
At Lusis Payments, technology is our passion, and our clients are the heart of our business. We live and breathe payments platform technology, and we’re committed to maintaining our position as a market leader. Beyond payments, though, we’re striving to pave the way for new technologies that will enhance our current solutions and lay the foundation for future solutions that will take the industry by storm. Here’s what our team at Lusis Payments was up to in 2018, as well as a look at what’s ahead in 2019.
Out and About
We’re devoted to our clients and technology, but that doesn’t mean we’re chained to our desks. Last year, you might have met us at a Tandem User Group (TUG) event near Atlanta, Georgia (ATUG), or in Mississauga, Ontario (CTUG). You may have seen us at Forum TELECOM Nancy at Université de Lorraine. Maybe you visited our booth at the HPE NonStop Technical Boot Camp in Burlingame, California.
Industry Game Changers Revised Payment Service Directive (PSD2)
PSD2 was implemented in 2018, and, per instructions from the European Union (EU), it requires banks to enable their customers to use third-party financial services vendors. This will shake up banks’ competition model. They’ll need to ensure that these third-party providers have access to their customers’ accounts through open Application Programming Interfaces (APIs), and third-parties will be able to build financial services on top of banks’ data and infrastructures. Competitive differentiation will become less about an account’s location and more about the services customers use. Read the full perspective by CEO, Philippe Préval at https://www.lusispayments.com/news/psd2-international-perspective
New ATM and POS Security Requirements
In August 2018, Lusis Payments released a report summarizing new security requirements introduced by PCI-PIN V3 that merge requirements from the PCI Security Standards Council (PCI SCC) and the Accredited Standards Committee X9 Inc (ASC X9) for a single, unified PIN Security Standard for payments stakeholders. Two evolutions — ASC X9 TR-31 and ANSI TR-34 — have emerged from the standards and will introduce a massive change in ATM and POS remote key management. X9 TR-31 is a method consistent with the requirements of the ANS X9.24-1 standard for the secure exchange of keys and other sensitive data between two devices that share a symmetric key exchange. ANSI TR-34 describes new rules to unify methods to build and store the original Terminal Master Key in ATMs and POS devices. Until now, each vendor has used its own methods to achieve this initialization. ANSI TR-34 specifies a method consistent with the requirements of ANSI X9.24-2. Read the complete Lusis article at: https://www.lusispayments.com/news/update-security-pci-pin-v3-tr34-tr31
The Future of ATMs
Last May, we reported about new technologies that will affect the future of ATMs. Banks have two new initiatives on the table for cash delivery:
The ATMs of the Future
Meanwhile, ATM vendors are introducing ways to expand typical ATM capabilities and help banks reduce labor costs without degrading the customer experience. These advanced banking machines provide assisted service hardware solutions similar to traditional ATMs, but additional software and hardware enable integrated two-way video conferencing with live tellers stationed at remote locations. The machines perform operations typically provided by live tellers, and they can serve as stand-alone video teller solutions or replace a portion of live teller agents inside branches. And the best part? Customers interact face to face with live tellers, so they still have the experience of personal interactions. Read the full perspective by CEO, Philippe Préval at: https://www.lusispayments.com/news/the-future-of-atms-new-initiatives-from-vendors-and-banks
Our Goal: Pave the Way for the Future of Payments Tech TANGO Cloud Achieves 10,000 Transactions Per Second
Our clients’ customers continually seek better, faster and more accessible financial services and products, so Lusis Payments is dedicated to continuous improvement and providing the industry-leading payments technology solutions. Recently, we conducted another proof-of-concept (POC) benchmark on TANGO, our cloud-based transaction processing engine for mission-critical, 24x7 payments.
The POC showcases the TANGO architecture’s ability in a cloud environment. Completion of the benchmark validated that TANGO not only meets but also exceeds the 10,000 TPS mark with a constant response time. In our next POC, Lusis Payments will demonstrate that TANGO can exceed the 20,000 TPS mark in a standard cloud environment, and our third POC will benchmark TANGO in an HPE NonStop Cloud environment.
Revolutionizing the FinTech Industry With Artificial Intelligence
Lusis Payments is at the top of our game when it comes to payments systems, and, while we remain dedicated to continuous improvement of our platforms, we’ve also been investing and delving deeper into other technologies to enhance our current solutions. Even more exciting is how these new technologies could potentially diversify future Lusis Payments solutions and position us as not only a market leader but also a market disrupter.
In 2017, we launched a dedicated AI department, which leads our AI-based trading strategies, other FinTech solutions, and global retail payments software and services. Our central focus is on deep learning, which offers extraordinary power and flexibility for problem resolution. Today, we offer a broad range of FinTech solutions and services for capital markets.
But the applications for artificial intelligence and data science are infinite; in many ways, the tech industry has only begun to tap in to their potential. AI includes tools like machine learning, natural language processing and deep neural networks — tech that’s improving efficiency in almost every industry. One area where AI has more recently become a disruptor is in capital market applications. According to a Gartner survey, organizations in the banking and investment services industry are more likely to actively invest in machine learning and virtual customer assistants than those in other industries.
Lusis Payments has spent five years using data science approaches to analyze and improve our FX trading platform because we see great potential for AI to change end-to-end trading, automatic trading and portfolio management. In fact, our experts predict that AI will influence every aspect of capital market applications.
Happy New Year from all of us at Lusis Payments. We hope you’ll check back to learn about our advancements and initiatives in 2019.
The latest company and industry news from Lusis Payments.