Search

Your search keyword '"Grossmann, Vasco"' showing total 17 results
Start Over Author "Grossmann, Vasco"
17 results on '"Grossmann, Vasco"'

1. Annotating Ambiguous Images: General Annotation Strategy for High-Quality Data with Real-World Biomedical Validation

Author

Schmarje, Lars, Grossmann, Vasco, Zelenka, Claudius, Brünger, Johannes, and Koch, Reinhard

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

In the field of image classification, existing methods often struggle with biased or ambiguous data, a prevalent issue in real-world scenarios. Current strategies, including semi-supervised learning and class blending, offer partial solutions but lack a definitive resolution. Addressing this gap, our paper introduces a novel strategy for generating high-quality labels in challenging datasets. Central to our approach is a clearly designed flowchart, based on a broad literature review, which enables the creation of reliable labels. We validate our methodology through a rigorous real-world test case in the biomedical field, specifically in deducing height reduction from vertebral imaging. Our empirical study, leveraging over 250,000 annotations, demonstrates the effectiveness of our strategies decisions compared to their alternatives., Comment: Accepted at ICLR 2024, DMLR Workshop

Published
2023

2. Label Smarter, Not Harder: CleverLabel for Faster Annotation of Ambiguous Image Classification with Higher Quality

Author

Schmarje, Lars, Grossmann, Vasco, Michels, Tim, Nazarenus, Jakob, Santarossa, Monty, Zelenka, Claudius, and Koch, Reinhard

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

High-quality data is crucial for the success of machine learning, but labeling large datasets is often a time-consuming and costly process. While semi-supervised learning can help mitigate the need for labeled data, label quality remains an open issue due to ambiguity and disagreement among annotators. Thus, we use proposal-guided annotations as one option which leads to more consistency between annotators. However, proposing a label increases the probability of the annotators deciding in favor of this specific label. This introduces a bias which we can simulate and remove. We propose a new method CleverLabel for Cost-effective LabEling using Validated proposal-guidEd annotations and Repaired LABELs. CleverLabel can reduce labeling costs by up to 30.0%, while achieving a relative improvement in Kullback-Leibler divergence of up to 29.8% compared to the previous state-of-the-art on a multi-domain real-world image classification benchmark. CleverLabel offers a novel solution to the challenge of efficiently labeling large datasets while also improving the label quality.

Published
2023

3. Beyond Hard Labels: Investigating data label distributions

Author

Grossmann, Vasco, Schmarje, Lars, and Koch, Reinhard

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

High-quality data is a key aspect of modern machine learning. However, labels generated by humans suffer from issues like label noise and class ambiguities. We raise the question of whether hard labels are sufficient to represent the underlying ground truth distribution in the presence of these inherent imprecision. Therefore, we compare the disparity of learning with hard and soft labels quantitatively and qualitatively for a synthetic and a real-world dataset. We show that the application of soft labels leads to improved performance and yields a more regular structure of the internal feature space., Comment: https://icml.cc/virtual/2022/workshop/13477

Published
2022

4. Is one annotation enough? A data-centric image classification benchmark for noisy and ambiguous label estimation

Author

Schmarje, Lars, Grossmann, Vasco, Zelenka, Claudius, Dippel, Sabine, Kiko, Rainer, Oszust, Mariusz, Pastell, Matti, Stracke, Jenny, Valros, Anna, Volkmann, Nina, and Koch, Reinhard

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

High-quality data is necessary for modern machine learning. However, the acquisition of such data is difficult due to noisy and ambiguous annotations of humans. The aggregation of such annotations to determine the label of an image leads to a lower data quality. We propose a data-centric image classification benchmark with ten real-world datasets and multiple annotations per image to allow researchers to investigate and quantify the impact of such data quality issues. With the benchmark we can study the impact of annotation costs and (semi-)supervised methods on the data quality for image classification by applying a novel methodology to a range of different algorithms and diverse datasets. Our benchmark uses a two-phase approach via a data label improvement method in the first phase and a fixed evaluation model in the second phase. Thereby, we give a measure for the relation between the input labeling effort and the performance of (semi-)supervised algorithms to enable a deeper insight into how labels should be created for effective model training. Across thousands of experiments, we show that one annotation is not enough and that the inclusion of multiple annotations allows for a better approximation of the real underlying class distribution. We identify that hard labels can not capture the ambiguity of the data and this might lead to the common issue of overconfident models. Based on the presented datasets, benchmarked methods, and analysis, we create multiple research opportunities for the future directed at the improvement of label noise estimation approaches, data annotation schemes, realistic (semi-)supervised learning, or more reliable image collection., Comment: Accepted at NeurIPS 2022, Benchmark and Dataset Track, Code and Link to data available at https://github.com/Emprime/dcic

Published
2022

6. Gemini connector

Author

Grossmann, Vasco, primary, Nakath, David, additional, Koch, Reinhard, additional, and Köser, Kevin, additional

Published
2022
Full Text
View/download PDF

8. Digital twinning in the ocean - chanllenges in multimodal sensing and multiscale fusion based on faithful visual models

Author

Grossmann, Vasco, Nakath, David, Urlaub, Morelia, Oppelt, Natascha, Koch, Reinhard, Köser, Kevin, Grossmann, Vasco, Nakath, David, Urlaub, Morelia, Oppelt, Natascha, Koch, Reinhard, and Köser, Kevin

Abstract

In engineering, machines are typically built after a careful conception and design process: All components of a system, their roles and the interaction between them is well understood, and often even digital models of the system exist before the actual hardware is built. This enables simulations and even feedback loops between the real-world system and a digital model, leading to a digital twin that allows better testing, prediction and understanding of complex effects. On the contrary, in Earth sciences, and particularly in ocean sciences, models exist only for certain aspects of the real world, of certain processes and of some interactions and dependencies between different “components” of the ocean. These individual models cover large temporal (seconds to millions of years) and spatial (millimetres to thousands of kilometres) scales, a variety of field data underpin them, and their results are represented in many different ways. A key to enabling digital twins in the oceans is fusion at different levels, in particular, fusion of data sources and modalities, fusion over different scales and fusion of differing representations. We outline these challenges and exemplify different envisioned digital twins employed in the oceans involving remote sensing, underwater photogrammetry and computer vision, focusing on optical aspects of the digital twinning process. In particular, we look at the holistic sensing scenarios of optical properties in coastal waters as well as seafloor dynamics at volcanic slopes and discuss road blockers for digital twins as well as potential solutions to increase and widen the use of digital twins.

Published
2022
Full Text
View/download PDF

9. Mean Reversion-orientiertes Swing Trading

Author

Grossmann, Vasco

Subjects
004 Data processing and computer science, Welfare economics, Economics, data science
Abstract

In dieser Dissertation wird ein Handelssystem beschrieben und evaluiert, welches rucklaufige Tendenzen nach Uberreaktionen in Finanzmarkten analysiert und ausnutzt. Hierfur werden Fehlbewertungen uber ein neues Verfahren stochastisch modelliert, um den tatsachlichen Wert eines Finanzprodukts von irrationalen Kursbewegungen zu separieren – die Analyse von Realmarktdaten offenbart langfristige Gleichgewichtsniveaus, zu dessen Ruckkehr untersuchte Kursprozesse tendieren. Verzogerte Wechselwirkungen zwischen unterschiedlichen Finanzprodukten helfen zusatzlich, Uberreaktionen zu erkennen. Erkenntnisse der Martingal- und Portfoliotheorie werden bei der Definition des Handelssystems mit dem Ziel kombiniert, identifizierte Fehlbewertungen in gleichmasige Gewinne zu uberfuhren. Auf diese Weise wird ein nachhaltiger und sicherheitsorientierter Vermogensaufbau begunstigt. Um moglichst prazise Aussagen uber den Erfolg des Handelssystems treffen zu konnen, werden Simulationen auf kunstlichen und realen Kursdaten untersucht. Ausgewahlte Performance- und Risikomase werden dabei einerseits modellbasiert ausgewertet, um den Einfluss einzelner Zeitreiheneigenschaften akkurat bestimmen zu konnen. Andererseits werden Realmarktdaten genutzt, um einen unverfalschten Blick uber den tatsachlichen Erfolg der theoretischen Uberlegungen zu ermoglichen. Fur die Evaluation werden Finanzmarktdaten von Aktien, Rohstoffen und Wechselkursen im Bereich von 1968 bis 2018 untersucht. Die abschliesende finanzwirtschaftliche Auswertung offenbart an einem umfassenden Beispiel, dass das Handelssystem in der Dekade von 2008 bis 2018 stetige Gewinne bei einem moderaten Risiko generiert hatte.%%%%Mean-reversion is a fundamental term in finance market analysis that refers to the assumption of an equilibrium in asset prices that is superimposed with overreactions. This dissertation describes and evaluates a trading system that exploits mean-reverse price tendencies in finance markets. Under the assumption of mispricings, a stochastic model is developed to separate fair price levels from irrational price movements. Analysis results reveal long-term balances in the probed market data. Delayed interdependencies between asset classes are utilized to improve the identification of overreactions. Methods of the martingale and portfolio theory are combined to convert these mispricings into steady returns with the objective to depict a sustainable and security-oriented trading system. Simulations on artifical and real quote data are combined to generate precise statements over the success of the trading system. On the one hand, performance and risk measures are evaluated in a model-based approach to quantify the influence of crucial time series properties. On the other hand, results on real market data allow unbiased performance and risk statements. The evaluation is based on finance market data of stocks, resources and foreign exchanges in the period of 1968 and 2018. Next to an isolated analysis of trading results for single instruments, the performance…

Published
2019
Full Text
View/download PDF

10. Swing Trading in Mean-reverting Markets

Author

Grossmann, Vasco, Prof. Dr. Manfred Schimmler, and Prof. Dr. Malte Braack

Subjects
Faculty of Engineering, Technische Fakultät, Technical Analysis of the Financial Markets, Time Series Analysis, Mean Reversion, Cointegration, ddc:004, Technische Finanzmarktanalyse, Zeitreihenanalyse, Mean Reversion, Kointegration
Abstract

In dieser Dissertation wird ein Handelssystem beschrieben und evaluiert, welches rückläufige Tendenzen nach Überreaktionen in Finanzmärkten analysiert und ausnutzt. Hierfür werden Fehlbewertungen über ein neues Verfahren stochastisch modelliert, um den tatsächlichen Wert eines Finanzprodukts von irrationalen Kursbewegungen zu separieren -- die Analyse von Realmarktdaten offenbart langfristige Gleichgewichtsniveaus, zu dessen Rückkehr untersuchte Kursprozesse tendieren. Verzögerte Wechselwirkungen zwischen unterschiedlichen Finanzprodukten helfen zusätzlich, Überreaktionen zu erkennen. Erkenntnisse der Martingal- und Portfoliotheorie werden bei der Definition des Handelssystems mit dem Ziel kombiniert, identifizierte Fehlbewertungen in gleichmäßige Gewinne zu überführen. Auf diese Weise wird ein nachhaltiger und sicherheitsorientierter Vermögensaufbau begünstigt. Um möglichst präzise Aussagen über den Erfolg des Handelssystems treffen zu können, werden Simulationen auf künstlichen und realen Kursdaten untersucht. Ausgewählte Performance- und Risikomaße werden dabei einerseits modellbasiert ausgewertet, um den Einfluss einzelner Zeitreiheneigenschaften akkurat bestimmen zu können. Andererseits werden Realmarktdaten genutzt, um einen unverfälschten Blick über den tatsächlichen Erfolg der theoretischen Überlegungen zu ermöglichen. Für die Evaluation werden Finanzmarktdaten von Aktien, Rohstoffen und Wechselkursen im Bereich von 1968 bis 2018 untersucht. Die abschließende finanzwirtschaftliche Auswertung offenbart an einem umfassenden Beispiel, dass das Handelssystem in der Dekade von 2008 bis 2018 stetige Gewinne bei einem moderaten Risiko generiert hätte. Mean-reversion is a fundamental term in finance market analysis that refers to the assumption of an equilibrium in asset prices that is superimposed with overreactions. This dissertation describes and evaluates a trading system that exploits mean-reverse price tendencies in finance markets. Under the assumption of mispricings, a stochastic model is developed to separate fair price levels from irrational price movements. Analysis results reveal long-term balances in the probed market data. Delayed interdependencies between asset classes are utilized to improve the identification of overreactions. Methods of the martingale and portfolio theory are combined to convert these mispricings into steady returns with the objective to depict a sustainable and security-oriented trading system. Simulations on artifical and real quote data are combined to generate precise statements over the success of the trading system. On the one hand, performance and risk measures are evaluated in a model-based approach to quantify the influence of crucial time series properties. On the other hand, results on real market data allow unbiased performance and risk statements. The evaluation is based on finance market data of stocks, resources and foreign exchanges in the period of 1968 and 2018. Next to an isolated analysis of trading results for single instruments, the performance of a portfolio of mixed asset classes is reviewed. A final evaluation from 2008 to 2018 depict steady returns over the whole decade. As a diversified asset composition smoothes equity requirements, the examined portfolio succeeds in a further risk reduction.

Published
2018

11. Mean Reversion-orientiertes Swing Trading: Technische Finanzmarktanalyse zur Optimierungeines Martingal-Handelssystems

Author

Grossmann, Vasco, Schimmler, Manfred, and Braack, Malte

Subjects
Mean Reversion, Technical Analysis of the Financial Markets, doctoral thesis, Abschlussarbeit, Kointegration, Cointegration, ddc:0XX, Technische Finanzmarktanalyse, Zeitreihenanalyse, ddc:004, Time Series Analysis
Abstract

In dieser Dissertation wird ein Handelssystem beschrieben und evaluiert, welches rückläufige Tendenzen nach Überreaktionen in Finanzmärkten analysiert und ausnutzt. Hierfür werden Fehlbewertungen über ein neues Verfahren stochastisch modelliert, um den tatsächlichen Wert eines Finanzprodukts von irrationalen Kursbewegungen zu separieren -- die Analyse von Realmarktdaten offenbart langfristige Gleichgewichtsniveaus, zu dessen Rückkehr untersuchte Kursprozesse tendieren. Verzögerte Wechselwirkungen zwischen unterschiedlichen Finanzprodukten helfen zusätzlich, Überreaktionen zu erkennen. Erkenntnisse der Martingal- und Portfoliotheorie werden bei der Definition des Handelssystems mit dem Ziel kombiniert, identifizierte Fehlbewertungen in gleichmäßige Gewinne zu überführen. Auf diese Weise wird ein nachhaltiger und sicherheitsorientierter Vermögensaufbau begünstigt. Um möglichst präzise Aussagen über den Erfolg des Handelssystems treffen zu können, werden Simulationen auf künstlichen und realen Kursdaten untersucht. Ausgewählte Performance- und Risikomaße werden dabei einerseits modellbasiert ausgewertet, um den Einfluss einzelner Zeitreiheneigenschaften akkurat bestimmen zu können. Andererseits werden Realmarktdaten genutzt, um einen unverfälschten Blick über den tatsächlichen Erfolg der theoretischen Überlegungen zu ermöglichen. Für die Evaluation werden Finanzmarktdaten von Aktien, Rohstoffen und Wechselkursen im Bereich von 1968 bis 2018 untersucht. Die abschließende finanzwirtschaftliche Auswertung offenbart an einem umfassenden Beispiel, dass das Handelssystem in der Dekade von 2008 bis 2018 stetige Gewinne bei einem moderaten Risiko generiert hätte. Mean-reversion is a fundamental term in finance market analysis that refers to the assumption of an equilibrium in asset prices that is superimposed with overreactions. This dissertation describes and evaluates a trading system that exploits mean-reverse price tendencies in finance markets. Under the assumption of mispricings, a stochastic model is developed to separate fair price levels from irrational price movements. Analysis results reveal long-term balances in the probed market data. Delayed interdependencies between asset classes are utilized to improve the identification of overreactions. Methods of the martingale and portfolio theory are combined to convert these mispricings into steady returns with the objective to depict a sustainable and security-oriented trading system. Simulations on artifical and real quote data are combined to generate precise statements over the success of the trading system. On the one hand, performance and risk measures are evaluated in a model-based approach to quantify the influence of crucial time series properties. On the other hand, results on real market data allow unbiased performance and risk statements. The evaluation is based on finance market data of stocks, resources and foreign exchanges in the period of 1968 and 2018. Next to an isolated analysis of trading results for single instruments, the performance of a portfolio of mixed asset classes is reviewed. A final evaluation from 2008 to 2018 depict steady returns over the whole decade. As a diversified asset composition smoothes equity requirements, the examined portfolio succeeds in a further risk reduction.

Published
2018

Catalog

Books, media, physical & digital resources
See catalog results