Spelling suggestions: "subject:"maschinelles lernen"" "subject:"maschinellen lernen""
171 |
VOCUS : a visual attention system for object detection and goal-directed search /Frintrop, Simone. January 1900 (has links)
Thesis (Ph.D.)--University of Bonn, Germany. / Includes bibliographical references and index. Also issued online.
|
172 |
New Paradigms for Automated Classification of PotteryHörr, Christian, Lindinger, Elisabeth, Brunnett, Guido 14 September 2009 (has links) (PDF)
This paper describes how feature extraction on ancient pottery can be combined with recent developments in artificial intelligence to draw up an automated, but still flexible classification system. These features include for instance several dimensions of the vessel's body, ratios thereof, an abstract representation of the overall shape, the shape of vessel segments and the number and type of attachments such as handles, lugs and feet. While most traditional approaches to classification are based on statistical analysis or the search for fuzzy clusters in high-dimensional spaces, we apply machine learning techniques, such as decision tree algorithms and neural networks. These methods allow for an objective and reproducible classification process. Conclusions about the "typability" of data, the evolution of types and the diagnostic attributes of the types themselves can be drawn as well.
|
173 |
Applications of Advanced Analytics to the Promotion of Freemium GoodsRunge, Julian 06 October 2020 (has links)
“Freemium” (Free + Premium) hat sich zu einem führenden Preismodell für digitale Güter entwickelt. Dabei kann die Basisversion eines Produkts, z.B. von Handy-Applikationen (“Apps”), unbegrenzt kostenlos genutzt werden und Firmen bieten Premium-Erweiterungen gegen Bezahlung an. Konsumenten haben in 2018 194 Milliarden mal Apps heruntergeladen und 101 Milliarden US-Dollar für In-App-Einkäufe ausgegeben. Beinahe 80% des Umsatzes auf App-Stores wird dabei durch Handyspiele generiert. 2,4 Milliarden Menschen haben in 2019 Handyspiele gespielt, was der Hälfte aller App-Nutzer im gleichen Zeitraum entspricht.
Die Hauptthese dieser Dissertation ist, dass preisreduzierende Sonderangebote von großer Wichtigkeit für das Vermarkten von Freemium-Gütern sind: Obwohl Freemium bereits eine extreme Preis-Reduktion darstellt, indem es ein Produkt Konsumenten kostenlos zum Ausprobieren zur Verfügung stellt, können demnach Firmen durch die Nutzung weiterer Sonderangebotstaktiken höhere Profite generieren. Die Arbeit postuliert weiter (und beweist dies empirisch), dass lange angenommene Risiken in der Nutzung von Sonderangeboten, die vor allem bei klassischen Konsumgütern etabliert wurden, im Freemium-Bereich in dieser Form nicht zutreffen. Diese Perspektive entwickelt und vertieft der Autor über vier individuelle Papiere, die zusammen mit einer einleitenden Zusammenfassung die fünf Kapitel dieser Dissertation ausmachen.
Die vorliegende Arbeit ist empirischer Natur und wendet “Advanced Analytics”, insbesondere Feldexperimente und maschinelles Lernen, in Zusammenarbeit mit Firmen an. Als repräsentativer Forschungsgrund dienen dabei Freemium-Handyspiele, in denen Firmen detaillierte Daten über Interaktionen mit Kunden sammeln. Anhand dieser Daten leitet der Autor neue Kenntnisse über Kundenverhalten ab und entwickelt Entscheidungsunterstützungssysteme, die es Firmen ermöglichen, höhere Gewinne beim Verkauf von Freemium-Gütern zu erzielen. / “Freemium” (free + premium) has become a workhorse pricing model in the digital economy: A basic version of a product or service, e.g., mobile applications (“apps”), can be used for free in perpetuity and premium upgrades are available against payment of a fee. Consumers downloaded apps 194 billion times in 2018 and spent $101 billion on in-app purchases in the same time period. Accounting for almost 80% of that revenue, gaming in particular has seen an unparalleled expansion of demand. It is estimated that 50% of mobile app users play games regularly and that a global total of 2.4 billion people will play mobile games in 2019.
The core thesis of this dissertation is that promotions are essential to the marketing of freemium goods such as mobile apps and games. While freemium already represents a promotional pricing tactic in using a zero price for free sampling, the author conjectures that firms can operate their freemium offerings more profitably by using further promotional tactics, especially targeted and personalized promotions, to sell premium upgrades. The author also argues (and shows) that widespread concerns around the use of promotions, particularly developed in the setting of consumer packaged goods, do not apply in the same way in this setting. This thinking is qualified and developed across four chapters that represent individual papers after providing an introduction to the work in the first chapter.
The work is empirical in nature and applies advanced analytics, in particular field experimentation and machine learning, in collaboration with firms. As representative of the freemium app economy, the collaborating firms observe dense user data that enable the author to both derive insights on consumer behavior that extend existing conceptual thinking in the field of marketing and to devise decision support and expert systems that allow firms to operate more profitably in this setting.
|
174 |
Logistic Regression for Prospectivity ModelingKost, Samuel 02 December 2020 (has links)
The thesis proposes a method for automated model selection using a logistic regression model in the context of prospectivity modeling, i.e. the exploration of minearlisations. This kind of data is characterized by a rare positive event and a large dataset. We adapted and combined the two statistical measures Wald statistic and Bayes' information criterion making it suitable for the processing of large data and a high number of variables that emerge in the nonlinear setting of logistic regression.
The obtained models of our suggested method are parsimonious allowing for an interpretation and information gain. The advantages of our method are shown by comparing it to another model selection method and to arti cial neural networks on several datasets. Furthermore we introduced a possibility to induce spatial dependencies which are important in such geological settings.
|
175 |
Simulations and data-based models for electrical conductivities of graphene nanolaminatesRothe, Tom 13 August 2021 (has links)
Graphene-based conductor materials (GCMs) consist of stacked and decoupled layers of graphene flakes and could potentially transfer graphene’s outstanding material properties like its exceptional electrical conductivity to the macro scale, where alternatives to the heavy and expensive metallic conductors are desperately needed. To reach super-metallic conductivity however, a systematic electrical conductivity optimization regarding the structural and physical input parameters is required. Here, a new trend in the field of process and material optimization are data-based models which utilize data science methods to quickly identify and abstract information and relationships from the available data. In this work such data-based models for the conductivity of a real GCM thin-film sample are build on data generated with an especially improved and extended version of the network simulation approach by Rizzi et al. [1, 2, 3]. Appropriate methods to create data-based models for GCMs are thereby introduced and typical challenges during the modelling process are addressed, so that data-based models for other properties of GCMs can be easily created as soon as sufficient data is accessible. Combined with experimental measurements by Slawig et al. [4] the created data-based models allow for a coherent and comprehensive description of the thin-films’
electrical parameters across several length scales.:List of Figures
List of Tables
Symbol Directory
List of Abbreviations
1 Introduction
2 Simulation approaches for graphene-based conductor materials
2.1 Traditional simulation approaches for GCMs
2.1.1 Analytical model for GCMs
2.1.2 Finite element method simulations for GCMs
2.2 A network simulation approach for GCMs
2.2.1 Geometry generation
2.2.2 Electrical network creation
2.2.3 Contact and probe setting
2.2.4 Conductivity computation
2.2.5 Results obtained with the network simulation approach
2.3 An improved implementation for the network simulation
2.3.1 Rizzi’s implementation of the network simulation approach
2.3.2 An network simulation tool for parameter studies
2.3.3 Extending the network simulation approach for anisotropy investigations and multilayer flakes
3 Data-based material modelling
3.1 Introduction to data-based modelling
3.2 Data-based modelling in material science
3.3 Interpretability of data-based models
3.4 The data-based modelling process
3.4.1 Preliminary considerations
3.4.2 Data acquisition
3.4.3 Preprocessing the data
3.4.4 Partitioning the dataset
3.4.5 Training the model
3.4.6 Model evaluation
3.4.7 Real-world applications
3.5 Regression estimators
3.5.1 Mathematical introduction to regression
3.5.2 Regularization and ridge regression
3.5.3 Support Vector Regression
3.5.4 Introducing non-linearity through kernels
4 Data-based models for a real GCM thin-film
4.1 Experimental measurements
4.2 Simulation procedure
4.3 Data generation
4.4 Creating data-based models
4.4.1 Quadlinear interpolation as benchmark model
4.4.2 KR, KRR and SVR
4.4.3 Enlarging the dataset
4.4.4 KR, KRR and SVR on the enlarged training dataset
4.5 Application to the GCM sample
5 Conclusion and Outlook
5.1 Conclusion
5.2 Outlook
Acknowledgements
Statement of Authorship
|
176 |
Schnelle Identifizierung von oralen Actinomyces-Arten des subgingivalen Biofilms mittels MALDI-TOF-MSBorgmann, Toralf Harald 10 November 2015 (has links)
Aktinomyzeten sind ein Teil der residenten Flora des menschlichen Verdauungstraktes, des Urogenitalsystems und der Haut. Die zeitraubende Isolation und Identifikation der Aktinomyzeten durch konventionelle Methoden stellt sich häufig als sehr schwierig dar. In den letzten Jahren hat sich jedoch die Matrix-unterstützte Laser-Desorption/Ionisation-Flugzeit-Massenspektrometrie (MALDI-TOF-MS) als Alternative zu etablierten Verfahren entwickelt und stellt heutzutage eine schnelle und simple Methode zur Bakterienidentifikation dar. Unsere Studie untersucht den Nutzen dieser Methode für eine schnelle und zuverlässige Identifizierung von oralen Aktinomyzeten, die aus dem subgingivalen Biofilm parodontal erkrankter Patienten isoliert wurden. In dieser Studie wurden elf verschiedene Referenzstämme aus den Stammsammlungen ATCC und DSMZ und 674 klinische Stämme untersucht. Alle Stämme wurden durch biochemische Methoden vorab identifiziert und anschließend ausgehend von den erhobenen MALDI-TOF-MS-Daten durch Ähnlichkeitsanalysen und Klassifikationsmethoden identifiziert und klassifiziert. Der Genotyp der Referenzstämme und von 232 klinischen Stämmen wurde durch Sequenzierung der 16S rDNA bestimmt. Die Sequenzierung bestätigte die Identifizierung der Referenzstämme. Diese und die zweifelsfrei durch 16S rDNA Sequenzierung identifizierten Aktinomyzeten wurden verwendet, um eine MALDI-TOF-MS-Datenbank zu erstellen. Methoden der Klassifikation wurden angewandt, um eine Differenzierung und Identifikation zu ermöglichen. Unsere Ergebnisse zeigen, dass eine Kombination aus Datenerhebung mittels MALDI-TOF-MS und deren Verarbeitung mittels SVM-Algorithmen eine gute Möglichkeit für die Identifikation und Differenzierung von oralen Aktinomyzeten darstellt.
|
177 |
Time Dynamic Topic ModelsJähnichen, Patrick 22 March 2016 (has links)
Information extraction from large corpora can be a useful tool for many applications in industry and academia. For instance, political communication science has just recently begun to use the opportunities that come with the availability of massive amounts of information available through the Internet and the computational tools that natural language processing can provide. We give a linguistically motivated interpretation of topic modeling, a state-of-the-art algorithm for extracting latent semantic sets of words from large text corpora, and extend this interpretation to cover issues and issue-cycles as theoretical constructs coming from political communication science. We build on a dynamic topic model, a model whose semantic sets of words are allowed to evolve over time governed by a Brownian motion stochastic process and apply a new form of analysis to its result. Generally this analysis is based on the notion of volatility as in the rate of change of stocks or derivatives known from econometrics. We claim that the rate of change of sets of semantically related words can be interpreted as issue-cycles, the word sets as describing the underlying issue. Generalizing over the existing work, we introduce dynamic topic models that are driven by general (Brownian motion is a special case of our model) Gaussian processes, a family of stochastic processes defined by the function that determines their covariance structure. We use the above assumption and apply a certain class of covariance functions to allow for an appropriate rate of change in word sets while preserving the semantic relatedness among words. Applying our findings to a large newspaper data set, the New York Times Annotated corpus (all articles between 1987 and 2007), we are able to identify sub-topics in time, \\\\textit{time-localized topics} and find patterns in their behavior over time. However, we have to drop the assumption of semantic relatedness over all available time for any one topic. Time-localized topics are consistent in themselves but do not necessarily share semantic meaning between each other. They can, however, be interpreted to capture the notion of issues and their behavior that of issue-cycles.
|
178 |
Spamerkennung mit Support Vector MachinesMöller, Manuel 22 June 2005 (has links)
Diese Arbeit zeigt ausgehend von einer Darstellung der theoretischen Grundlagen automatischer Textklassifikation, dass die aus der Statistical Learning Theory stammenden Support Vector Machines geeignet sind, zu einer präziseren Erkennung unerwünschter E-Mail-Werbung beizutragen. In einer Testumgebung mit einem Corpus von 20 000 E-Mails wurden Testläufe verschiedene Parameter der Vorverarbeitung und der Support Vector Machine automatisch evaluiert und grafisch visualisiert. Aufbauend darauf wird eine Erweiterung für die Open-Source-Software SpamAssassin beschrieben, die die vorhandenen Klassifikationsmechanismen um eine Klassifikation per Support Vector Machine erweitert.
|
179 |
Strategien zur Datenfusion beim Maschinellen LernenSchwalbe, Karsten, Groh, Alexander, Hertwig, Frank, Scheunert, Ulrich 25 November 2019 (has links)
Smarte Prüfsysteme werden ein Schlüsselbaustein zur Qualitätssicherung in der industriellen Fertigung und Produktion sein. Insbesondere trifft dies auf komplexe Prüf- und Bewertungsprozesse zu. In den letzten Jahren haben sich hierfür lernbasierte Verfahren als besonders vielversprechend herauskristallisiert. Ihr Einsatz geht in der Regel mit erheblichen Performanceverbesserungen gegenüber konventionellen, regel- bzw. geometriebasierten Methoden einher. Der Black-Box-Charakter dieser Algorithmen führt jedoch dazu, dass die Interpretationen der berechneten Prognosegüten kritisch zu hinterfragen sind. Das Vertrauen in die Ergebnisse von Algorithmen, die auf maschinellem Lernen basieren, kann erhöht werden, wenn verschiedene, voneinander unabhängige Verfahren zum Einsatz kommen. Hierbei sind Datenfusionsstrategien anzuwenden, um die Resultate der verschiedenen Methoden zu einem Endergebnis zusammenzufassen. Im Konferenzbeitrag werden, aufbauend auf einer kurzen Vorstellung wichtiger Ansätze zur Objektklassifikation, entsprechende Fusionsstrategien präsentiert und an einem Fallbeispiel evaluiert. Im Anschluss wird auf Basis der Ergebnisse das Potential der Datenfusion in Bezug auf das Maschinelle Lernen erörtert.
|
180 |
Radiomics risk modelling using machine learning algorithms for personalised radiation oncologyLeger, Stefan 18 June 2019 (has links)
One major objective in radiation oncology is the personalisation of cancer treatment. The implementation of this concept requires the identification of biomarkers, which precisely predict therapy outcome. Besides molecular characterisation of tumours, a new approach known as radiomics aims to characterise tumours using imaging data. In the context of the presented thesis, radiomics was established at OncoRay to improve the performance of imaging-based risk models. Two software-based frameworks were developed for image feature computation and risk model construction. A novel data-driven approach for the correction of intensity non-uniformity in magnetic resonance imaging data was evolved to improve image quality prior to feature computation. Further, different feature selection methods and machine learning algorithms for time-to-event survival data were evaluated to identify suitable algorithms for radiomics risk modelling. An improved model performance could be demonstrated using computed tomography data, which were acquired during the course of treatment. Subsequently tumour sub-volumes were analysed and it was shown that the tumour rim contains the most relevant prognostic information compared to the corresponding core. The incorporation of such spatial diversity information is a promising way to improve the performance of risk models.:1. Introduction
2. Theoretical background
2.1. Basic physical principles of image modalities
2.1.1. Computed tomography
2.1.2. Magnetic resonance imaging
2.2. Basic principles of survival analyses
2.2.1. Semi-parametric survival models
2.2.2. Full-parametric survival models
2.3. Radiomics risk modelling
2.3.1. Feature computation framework
2.3.2. Risk modelling framework
2.4. Performance assessments
2.5. Feature selection methods and machine learning algorithms
2.5.1. Feature selection methods
2.5.2. Machine learning algorithms
3. A physical correction model for automatic correction of intensity non-uniformity
in magnetic resonance imaging
3.1. Intensity non-uniformity correction methods
3.2. Physical correction model
3.2.1. Correction strategy and model definition
3.2.2. Model parameter constraints
3.3. Experiments
3.3.1. Phantom and simulated brain data set
3.3.2. Clinical brain data set
3.3.3. Abdominal data set
3.4. Summary and discussion
4. Comparison of feature selection methods and machine learning algorithms
for radiomics time-to-event survival models
4.1. Motivation
4.2. Patient cohort and experimental design
4.2.1. Characteristics of patient cohort
4.2.2. Experimental design
4.3. Results of feature selection methods and machine learning algorithms evaluation
4.4. Summary and discussion
5. Characterisation of tumour phenotype using computed tomography imaging
during treatment
5.1. Motivation
5.2. Patient cohort and experimental design
5.2.1. Characteristics of patient cohort
5.2.2. Experimental design
5.3. Results of computed tomography imaging during treatment
5.4. Summary and discussion
6. Tumour phenotype characterisation using tumour sub-volumes
6.1. Motivation
6.2. Patient cohort and experimental design
6.2.1. Characteristics of patient cohorts
6.2.2. Experimental design
6.3. Results of tumour sub-volumes evaluation
6.4. Summary and discussion
7. Summary and further perspectives
8. Zusammenfassung
|
Page generated in 0.1505 seconds