TY - JOUR A1 - Gosh Dastidar, Kanishka A1 - Caelen, Olivier A1 - Granitzer, Michael T1 - Machine learning methods for credit card fraud detection : a survey JF - IEEE Access N2 - The widespread adoption of online payments has been accompanied by a significant increase in fraudulent activities, resulting in billions of dollars in financial losses. As payment providers aim to tackle this with various preventive mechanisms, fraudsters also continuously evolve their methods to remain indistinguishable from genuine actors. This necessitates sophisticated fraud detection tools to supplement these security mechanisms. As the volume of transactions taking place per day is in the millions, relying solely on human investigation is expensive and ultimately unfeasible, leading to an emergence of research into data driven or statistical methods for fraud detection. Over the last decade, this research has evolved to tackle the various particularities of the domain. These include the skewed nature of the data, the evolving user and fraud behavior, and the learning representations of the context in which a transaction takes place. This work aims to provide the community with an in-depth overview of the different directions in which recent research on online fraud detection has focused. We develop a taxonomy of the domain based on these directions and organize our analysis accordingly. For each area, we focus on significant methodological advancements and highlight limitations or gaps in the current state-of-the-art solutions. Through our analysis, it emerges that one of the primary limiting factors that many researchers face is the lack of availability of high-quality credit card data. Therefore, we provide a first step in addressing this issue in the form of a data generation framework using generative adversarial networks (GANs). We hope that this survey serves as a foundation for researchers who want to address the multi-faceted problem of credit card fraud detection. KW - fraud detection KW - machine learning KW - neural networks KW - synthetic data Y1 - 2024 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:739-opus4-15796 VL - 2024 IS - 12 SP - 158939 EP - 158965 PB - IEEE CY - New York ER - TY - JOUR A1 - Caspari-Sadeghi, Sima T1 - Artificial Intelligence in Technology-Enhanced Assessment: A Survey of Machine Learning JF - Journal of Educational Technology Systems N2 - Intelligent assessment, the core of any AI-based educational technology, is defined as embedded, stealth and ubiquitous assessment which uses intelligent techniques to diagnose the current cognitive level, monitor dynamic progress, predict success and update students’ profiling continuously. It also uses various technologies, such as learning analytics, educational data mining, intelligent sensors, wearables and machine learning. This can be the key to Precision Education (PE): adaptive, tailored, individualized instruction and learning. This paper explores (a) the applications of Machine Learning (ML) in intelligent assessment, and (b) the use of deep learning models in ‘knowledge tracing and student modeling’. The paper concludes by discussing barriers involved in using state-of-the-art ML methods and some suggestions to unleash the power of data and ML to improve educational decision-making. KW - artificial intelligence KW - knowledge tracing KW - machine learning KW - technology-enhanced assessment (TEA) Y1 - 2022 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:739-opus4-11818 SN - 0047-2395 SN - 1541-3810 VL - 51 IS - 3 SP - 372 EP - 386 PB - SAGE Publications CY - Sage CA: Los Angeles, CA ER - TY - THES A1 - Wild, Matthias T1 - A statistical perspective on machine learning: three essays on improvement, application and interpretation of predictive algorithms N2 - In this thesis we address a series of topics at the interface between statistics and machine learning. This includes the improvement of predictive algorithms using statistical concepts, potential and challenges in the application of predictive algorithms in areas traditionally covered by statistical theory, and the interpretation of black-box algorithms by means of statistical hypothesis testing. KW - machine learning KW - hypothesis test KW - statistical learning KW - boosting KW - random forest Y1 - 2025 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:739-opus4-15311 ER - TY - JOUR A1 - Anagnostopoulos, Nikolaos Athanasios A1 - Teymuri, Benyamin A1 - Serati, Reza A1 - Rasti, Mehdi ED - Xie, Bin ED - Wang, Ning ED - Gu, Yi ED - Stefanidis, Angelos T1 - LP-MAB: Improving the Energy Efficiency of LoRaWAN Using a Reinforcement-Learning-Based Adaptive Configuration Algorithm JF - Sensors N2 - In the Internet of Things (IoT), Low-Power Wide-Area Networks (LPWANs) are designed to provide low energy consumption while maintaining a long communications’ range for End Devices (EDs). LoRa is a communication protocol that can cover a wide range with low energy consumption. To evaluate the efficiency of the LoRa Wide-Area Network (LoRaWAN), three criteria can be considered, namely, the Packet Delivery Rate (PDR), Energy Consumption (EC), and coverage area. A set of transmission parameters have to be configured to establish a communication link. These parameters can affect the data rate, noise resistance, receiver sensitivity, and EC. The Adaptive Data Rate (ADR) algorithm is a mechanism to configure the transmission parameters of EDs aiming to improve the PDR. Therefore, we introduce a new algorithm using the Multi-Armed Bandit (MAB) technique, to configure the EDs’ transmission parameters in a centralized manner on the Network Server (NS) side, while improving the EC, too. The performance of the proposed algorithm, the Low-Power Multi-Armed Bandit (LP-MAB), is evaluated through simulation results and is compared with other approaches in different scenarios. The simulation results indicate that the LP-MAB’s EC outperforms other algorithms while maintaining a relatively high PDR in various circumstances. KW - Internet of Things (IoT) KW - LoRaWAN KW - adaptive configuration KW - machine learning KW - reinforcement learning Y1 - 2023 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:739-opus4-11853 SN - 1424-8220 VL - 23 IS - 4 PB - MDPI CY - Basel, Switzerland ER - TY - THES A1 - Planche, Benjamin T1 - Bridging the Realism Gap for CAD-Based Visual Recognition N2 - Computer vision aims at developing algorithms to extract high-level information from images and videos. In the industry, for instance, such algorithms are applied to guide manufacturing robots, to visually monitor plants, or to assist human operators in recognizing specific components. Recent progress in computer vision has been dominated by deep artificial neural network, i.e., machine learning methods simulating the way that information flows in our biological brains, and the way that our neural networks adapt and learn from experience. For these methods to learn how to accurately perform complex visual tasks, large amounts of annotated images are needed. Collecting and labeling such domain-relevant training datasets is, however, a tedious—sometimes impossible—task. Therefore, it has become common practice to leverage pre-available three-dimensional (3D) models instead, to generate synthetic images for the recognition algorithms to be trained on. However, methods optimized over synthetic data usually suffer a significant performance drop when applied to real target images. This is due to the realism gap, i.e., the discrepancies between synthetic and real images (in terms of noise, clutter, etc.). In my work, three main directions were explored to bridge this gap. First, an innovative end-to-end framework is proposed to render realistic depth images from 3D models, as a growing number of solutions (especially in the industry) are utilizing low-cost depth cameras (e.g., Microsoft Kinect and Intel RealSense) for recognition tasks. Based on a thorough study of these devices and the different types of noise impairing them, the proposed framework simulates their inner mechanisms, comprehensively modeling vital factors such as sensor noise, material reflectance, surface geometry, etc. Able to simulate a wide panel of depth sensors and to quickly generate large datasets, this framework is used to train algorithms for various recognition tasks, consistently and significantly enhancing their performance compared to other state-of-the-art simulation tools. In some cases, however, relevant 2D or 3D object representations to generate synthetic samples are not available. Considering this different case of data scarcity, a solution is then proposed to incrementally build a representation of visual scenes from partial observations. Provided observations are localized from one to another based on their content and registered in a global memory with spatial properties. Simultaneously, this memory can be queried to render novel views of the scene. Furthermore, unobserved regions can be hallucinated in memory, in consistence with previous observations, hallucinations, and global priors. The efficacy of the proposed mnemonic and generative system, trainable end-to-end, is demonstrated on various 2D and 3D use-cases. Finally, an advanced convolutional neural network pipeline is introduced, tackling the realism gap from a novel angle. While most methods addressing this problem focus on bringing synthetic samples—or the knowledge acquired from them—closer to the real target domain, the proposed solution performs the opposite process, mapping unseen target images into controlled synthetic domains. The pre-processed samples can then be handed to downstream recognition methods, themselves purely trained on similar synthetic data, to greatly improve their accuracy. For each approach, a variety of qualitative and quantitative studies are detailed, providing successful comparisons to state-of-the-art methods. By proposing solutions to bridge the realism gap from either side, as well as a pipeline to improve the acquisition and generation of new visual content, this thesis provides a unique perspective on the challenges of data scarcity when building robust recognition systems. N2 - Die Computer Vision strebt an, Algorithmen zum Extrahieren hochwertiger Informationen von Bildern und Videos zu entwickeln. In der Industrie werden solche Algorithmen beispielsweise angewendet, um Fertigungsroboter zu steuern, um Betriebe visuell zu überwachen, oder um Mitarbeiter bei der Erkennung bestimmter Komponenten zu unterstützen. Die kürzlichen Fortschritte im Bereich Computer Vision wurden von tiefen künstlichen neuronalen Netzen dominiert. Diese Methoden des maschinelles Lernens (Machine Learning) simulieren die Art und Weise, in der die Information in unseren biologischen Gehirnen verarbeitet wird und in der unsere neuronale Netze sich anpassen und aus Erfahrung lernen. Damit diese Methoden zur genauen Ausführung komplexer visueller Aufgaben befähigt werden, müssen sie mit einer großen Anzahl von annotierten Bildern trainiert werden. Die Erhebung und Kennzeichnung entsprechender Trainingsdatensätze ist jedoch eine langwierige und manchmal sogar unmögliche Aufgabe. Deswegen ist es zur gängigen Praxis geworden, stattdessen die vorhandenen 3D-Modelle zur Generierung synthetischer Bilder einzusetzen, damit die Erkennungsalgorithmen mit Hilfe dieser Bilder trainiert werden. Allerdings, bei der Anwendung auf die realen Zielbilder, erleiden die Methoden, die durch synthetische Daten angepasst wurden, einen erheblichen Leistungsabfall. Dies geschieht aufgrund der Realismuslücke (Realism Gap), das heißt durch die Diskrepanzen zwischen synthetischen und realen Bildern (hinsichtlich von Rauschen, Störungen usw.). In meiner Arbeit wurden drei Hauptrichtungen untersucht, um diese Lücke zu schließen. Zuerst wird ein innovatives End-to-End-Framework vorgeschlagen, um realistische Tiefenbilder von 3D-Modellen zu rendern, denn immer mehr Lösungen (insbesondere in der Industrie) verwenden kostengünstige Tiefen-Kameras (z. B. Microsoft Kinect und Intel RealSense) für die Erkennungsaufgaben. Aufgrund einer gründlichen Untersuchung dieser Geräte und der verschiedenen Arten von Rauschen, die dem Aufnahmen beeinträchtigen, simuliert das vorgeschlagene Framework deren innere Mechanismen, indem Schlüsselfaktoren wie Sensorrauschen, Reflektionsgrade der Materialien, Oberflächengeometrie usw. umfassend modelliert werden. Dieses Framework ist in der Lage eine breite Palette von Tiefensensoren zu simulieren und schnell große Datensätze zu generieren. Dies wird eingesetzt, um die Algorithmen für verschiedene Erkennungsaufgaben zu trainieren und deren Leistung im Vergleich zu anderen hochmodernen Simulationsmethoden konsistent und erheblich zu verbessern. In manchen Fällen sind jedoch keine relevanten 2D- oder 3D-Objektdarstellungen zur Erzeugung von synthetischen Bildern verfügbar. Ausgehend von dieser Problematik des Datenmangels wurde eine Lösung vorgeschlagen, in der die Rekonstruktion von visuellen Szenen aus Teilbeobachtungen schrittweise durchgeführt wird. Die Bilder werden anhand ihres Inhalts in Bezug zueinander lokalisiert und in einer globalen Gedächtnisstruktur mit räumlichen Eigenschaften registriert Gleichzeitig kann dieses Gedächtnis abgerufen werden, um neuen Ansichten der Szene zu rendern. Darüber hinaus können bisher unbeobachtete Regionen in Übereinstimmung mit früheren Beobachtungen, Halluzinationen und globalen Vorwissen im Gedächtnis halluziniert werden. Die Wirksamkeit des vorgeschlagenen, durchgehend trainierbaren mnemonischen und generativen Systems, wird anhand von verschiedenen 2D- und 3D-Anwendungsfällen demonstriert. Schließlich wird eine auf Convolutional Neural Networks (CNNs) basierte weiter entwickelte Pipeline vorgestellt, die die Realismuslücke aus einem neuen Blickwinkel angeht. Während die meisten Methoden, die sich mit diesem Problem befassen, sich darauf konzentrieren, synthetische Datenproben (bzw. daraus erworbenes Wissen) näher an die echte/reale Zieldomäne zu bringen, führt die vorgeschlagene Lösung den umgekehrten Prozess durch, indem ungesehene Zielbilder in den kontrollierten synthetischen Domänen abgebildet werden. Die vorbehandelten Datenproben können dann für die nachgeschalteten Erkennungsalgorithmen übergeben werden, die selbst anhand der ähnlichen synthetischen Daten trainiert wurden, um deren Genauigkeit deutlich zu verbessern. Für jeden Ansatz werden verschiedene qualitative und quantitative Studien durchgeführt, um mit sie den neuesten Methoden zu vergleichen. Insgesamt werden in dieser Arbeit Methoden zur Überbrückung der Realismuslücke auf beiden Seiten sowie eine Lösung zur Verbesserung der Erfassung und Generierung neuer visueller Inhalte beschrieben. Daher bietet diese Dissertation eine neuartige Perspektive auf die Herausforderungen der Datenknappheit bei der Entwicklung robuster Erkennungssysteme. KW - computer vision KW - machine learning KW - domain adaptation KW - realism gap KW - visual understanding Y1 - 2020 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:739-opus4-8361 ER - TY - THES A1 - Wahl, Florian T1 - Methods for monitoring the human circadian rhythm in free-living N2 - Our internal clock, the circadian clock, determines at which time we have our best cognitive abilities, are physically strongest, and when we are tired. Circadian clock phase is influenced primarily through exposure to light. A direct pathway from the eyes to the suprachiasmatic nucleus, where the circadian clock resides, is used to synchronise the circadian clock to external light-dark cycles. In modern society, with the ability to work anywhere at anytime and a full social agenda, many struggle to keep internal and external clocks synchronised. Living against our circadian clock makes us less efficient and poses serious health impact, especially when exercised over a long period of time, e.g. in shift workers. Assessing circadian clock phase is a cumbersome and uncomfortable task. A common method, dim light melatonin onset testing, requires a series of eight saliva samples taken in hourly intervals while the subject stays in dim light condition from 5 hours before until 2 hours past their habitual bedtime. At the same time, sensor-rich smartphones have become widely available and wearable computing is on the rise. The hypothesis of this thesis is that smartphones and wearables can be used to record sensor data to monitor human circadian rhythms in free-living. To test this hypothesis, we conducted research on specialised wearable hardware and smartphones to record relevant data, and developed algorithms to monitor circadian clock phase in free-living. We first introduce our smart eyeglasses concept, which can be personalised to the wearers head and 3D-printed. Furthermore, hardware was integrated into the eyewear to recognise typical activities of daily living (ADLs). A light sensor integrated into the eyeglasses bridge was used to detect screen use. In addition to wearables, we also investigate if sleep-wake patterns can be revealed from smartphone context information. We introduce novel methods to detect sleep opportunity, which incorporate expert knowledge to filter and fuse classifier outputs. Furthermore, we estimate light exposure from smartphone sensor and weather in- formation. We applied the Kronauer model to compare the phase shift resulting from head light measurements, wrist measurements, and smartphone estimations. We found it was possible to monitor circadian phase shift from light estimation based on smartphone sensor and weather information with a weekly error of 32±17min, which outperformed wrist measurements in 11 out of 12 participants. Sleep could be detected from smartphone use with an onset error of 40±48 min and wake error of 42±57 min. Screen use could be detected smart eyeglasses with 0.9 ROC AUC for ambient light intensities below 200lux. Nine clusters of ADLs were distinguished using Gaussian mixture models with an average accuracy of 77%. In conclusion, a combination of the proposed smartphones and smart eyeglasses applications could support users in synchronising their circadian clock to the external clocks, thus living a healthier lifestyle. KW - context recognition KW - human circadian rhythm KW - machine learning KW - sleep timing KW - smart eyeglasses KW - Tagesrhythmus Y1 - 2020 U6 - http://nbn-resolving.de/urn/resolver.pl?urn:nbn:de:bvb:739-opus4-7607 ER -