Publication Server of Zuse Institute Berlin (ZIB)
Not a member yet
6648 research outputs found
Sort by
Evaluation of Model Bias Identification Approaches Based on Bayesian Inference and Applications to Digital Twins
No full textIn recent years, the use of simulation-based digital twins for monitoring and assessment of complex mechanical systems has greatly expanded. Their potential to increase the information obtained from limited data makes them an invaluable tool for a broad range of real-world applications. Nonetheless, there usually exists a discrepancy between the predicted response and the measurements of the system once built. One of the main contributors to this difference in addition to miscalibrated model parameters is the model error. Quantifying this socalled model bias (as well as proper values for the model parameters) is critical for the reliable performance of digital twins. Model bias identification is ultimately an inverse problem where information from measurements is used to update the original model. Bayesian formulations can tackle this task. Including the model bias as a parameter to be inferred enables the use of a Bayesian framework to obtain a probability distribution that represents the uncertainty between the measurements and the model. Simultaneously, this procedure can be combined with a classic parameter updating scheme to account for the trainable parameters in the original model.
This study evaluates the effectiveness of different model bias identification approaches based on Bayesian inference methods. This includes more classical approaches such as direct parameter estimation using MCMC in a Bayesian setup, as well as more recent proposals such as stat-FEM or orthogonal Gaussian Processes. Their potential use in digital twins, generalization capabilities, and computational cost is extensively analyzed
Using Multiple Reference Vectors and Objective Scaling in the Feasibility Pump
No full textThe Feasibility Pump (FP) is one of the best-known primal heuristics for mixed-integer programming (MIP): more than 15 papers suggested various modifications of all of its steps. So far, no variant considered information across multiple iterations, but all instead maintained the principle to optimize towards a single reference integer point. In this paper, we evaluate the usage of multiple reference vectors in all stages of the FP algorithm. In particular, we use LP-feasible vectors obtained during the main loop to tighten the variable domains before entering the computationally expensive enumeration stage. Moreover, we consider multiple integer reference vectors to explore further optimizing directions and introduce alternative objective scaling terms to balance the contributions of the distance functions and the original MIP objective. Our computational experiments demonstrate that the new method can improve performance on general MIP test sets. In detail, our modifications provide a 29.3% solution quality improvement and 4.0% running time improvement in an embedded setting, needing 16.0% fewer iterations over a large test set of MIP instances. In addition, the method’s success rate increases considerably within the first few iterations. In a standalone setting, we also observe a moderate performance improvement, which makes our version of FP suitable for the two main use-cases of the algorithm
Accurate reduced models for the pH oscillations in the urea-urease reaction confined to giant lipid vesicles
No full textThis theoretical study concerns a pH oscillator based on the urea-urease reaction confined to giant lipid vesicles. Under suitable conditions, differential transport of urea and hydrogen ion across the unilamellar vesicle membrane periodically resets the pH clock that switches the system from acid to basic, resulting in self-sustained oscillations. We analyse the structure of the phase flow and of the limit cycle, which controls the dynamics for giant vesicles and dominates the pronouncedly stochastic oscillations in small vesicles of submicrometer size. To this end, we derive reduced models, which are amenable to analytic treatments that are complemented by numerical solutions, and obtain the period and amplitude of the oscillations as well as the parameter domain, where oscillatory behavior persists. We show that the accuracy of these predictions is highly sensitive to the employed reduction scheme. In particular, we suggest an accurate two-variable model and show its equivalence to a three-variable model that admits an interpretation in terms of a chemical reaction network. The faithful modeling of a single pH oscillator appears crucial for rationalizing experiments and understanding communication of vesicles and synchronization of rhythms
Forward and Line-Based Cycle Bases for Periodic Timetabling
No full textThe optimization of periodic timetables is an indispensable planning task in public transport. Although the periodic event scheduling problem (PESP) provides an elegant mathematical formulation of the periodic timetabling problem that led to many insights for primal heuristics, it is notoriously hard to solve to optimality. One reason is that for the standard mixed-integer linear programming formulations, linear programming relaxations are weak and the integer variables are of pure technical nature and in general do not correlate with the objective value. While the first problem has been addressed by developing several families of cutting planes, we focus on the second aspect. We discuss integral forward cycle bases as a concept to compute improved dual bounds for PESP instances. To this end, we develop the theory of forward cycle bases on general digraphs. Specifically for the application of timetabling, we devise a generic procedure to construct line-based event-activity networks, and give a simple recipe for an integral forward cycle basis on such networks. Finally, we analyze the 16 railway instances of the benchmark library PESPlib, match them to the line-based structure and use forward cycle bases to compute better dual bounds for 14 out of the 16 instances
Faster exact solution of sparse MaxCut and QUBO problems
No full textThe maximum-cut problem is one of the fundamental problems in combinatorial optimization. With the advent of quantum computers, both the maximum-cut and the equivalent quadratic unconstrained binary optimization problem have experienced much interest in recent years. This article aims to advance the state of the art in the exact solution of both problems—by using mathematical programming techniques. The main focus lies on sparse problem instances, although also dense ones can be solved. We enhance several algorithmic components such as reduction techniques and cutting-plane separation algorithms, and combine them in an exact branch-and-cut solver. Furthermore, we provide a parallel implementation. The new solver is shown to significantly outperform existing state-of-the-art software for sparse maximum-cut and quadratic unconstrained binary optimization instances. Furthermore, we improve the best known bounds for several instances from the 7th DIMACS Challenge and the QPLIB, and solve some of them (for the first time) to optimality
Unsupervised Shape Correspondence Estimation for Anatomical Shapes
No full textThe concept of shape correspondence describes a relation between two or more shapes of the same class. It often consists of a mapping between points on semantically similar locations of all shapes. One possible application for shape correspondence in medicine is the automatic location of anatomical landmarks. Another popular application is the construction of statistical shape models. These models are an established way to represent geometric variation of anatomical shapes in a compact way. Possible applications range from the generation of shapes and reconstruction tasks to disease classification.
This thesis aims to investigate unsupervised methods that can be used to estimate such a correspondence on anatomical shapes. While most methods used in the medical domain focus on classical optimization algorithms to establish correspondence, the broader computer vision domain developed a versatile field of data-driven methods. Recently, the new shape model FlowSSM was introduced, which does not require predefined correspondences for training as it generates them itself. As the performance of the shape model is quite competitive, it is natural to assume that the generated correspondences are of high quality as well.
For this reason, we evaluate the quality of the correspondences generated by FlowSSM within this thesis. Furthermore, we modify the method by adding a second loss term that minimizes geodesic distortions. This is done to favor isometric deformations which can lead to better correspondences. We compare the results with two established methods from the medical domain, LDDMM and Meshmonk. Furthermore, we investigate the performance of a fourth method called Neuromoph. This data-driven method comes from the wider computer vision field and was not tested on anatomical data yet.
All methods are evaluated with a set of different metrics. This includes metrics to assess the quality of the resulting meshes, a sparse correspondence error on anatomical landmarks, and metrics to measure the quality of the resulting shape models. Furthermore, we test all methods on three datasets with different degrees of geometric variation, namely liver, distal femur and face.
We show that FlowSSM produces correspondences with state-of-the-art quality. Moreover, our modification further improved the quality of correspondences at a global level. Nevertheless, there is no clear ranking between all methods, as the results differ between metrics and datasets. Thereby, we can show that there are different qualities to a proper correspondence which are reflected in the different metrics. It is therefore strongly recommendable to choose a correspondence estimation method specifically for the problem at hand.Das Konzept der Formkorrespondenz zwischen 3D-Objekten einer Klasse beschreibt eine Beziehung zwischen den Instanzen (oft Punkten) der unterschiedlichen Objekten. Hierbei werden Punkte, die an semantisch gleichwertigen Orten liegen, miteinander in Verbindung gebracht. Eine mögliche Anwednung der Formkorrespondenz im medizinischen Bereich ist daher die automatisierte Lokalisierung von anatomischen Landmarken. Eine weitere Anwendung ist das Erstellen von statistischen Formmodellen. Mit diesen kann die geometrische Variation anatomischer Formen kompakt abgebildet werden. Medizinische Anwendungen reichen dabei von der einfachen Formgenerierung zu komplexeren Rekonstruktionsaufgaben und der Klassifizierung von gesunden und pathologischen Formen.
In dieser Arbeit werden unterschiedliche Methoden zur Erzeugung von Formkorrespondenzen untersucht. Die entsprechende Literatur im medizinischen Bereich verwendet hierzu meist Methoden, die das klassische Optimierungsproblem einer nichtrigiden Transformation lösen. Im Computer Vision Bereich wurden in den letzten Jahren auch einige datengetriebene Methoden zur Korrespondenzgenerierung veröffentlicht. Im letzten Jahr wurde außerdem die Methode FlowSSM zur Erstellung statistischer Formmodelle vorgestellt, die nicht auf korrespondierenden Oberflächen basiert, sondern diese selbst erzeugt. Da FlowSSM trotzdem konkurenzfähige Ergebnisse erzielt, ist naheliegend, dass auch die zugrundeliegenden, selbst generierten Korrespondenzen von hoher Qualität sind.
Innerhalb dieser Arbeit wird daher die Qualität der von FlowSSM erzeugten Korrespondenzen evaluiert. Außerdem wird die Methode um eine zusätzliche Kostenfunktion erweitert, die geod#tische Verzerrungen verhindern soll. Dadurch sollen nichtisometrische Deformationen vermieden werden, wodurch die Qualität der resultierenden Korrenspondenzen gesteigert werden kann. Die Ergebnisse von FlowSSM werden mit zwei etablierten Methoden aus dem medizinischen Bereich, LDDMM und Meshmonk, verglichen. Außerdem wird NeuroMorph, eine aktuelle, datengetriebene Methode aus dem Bereich des maschinellen Sehens getestet. Letztere wurde bisher noch nicht auf medizinischen Daten evaluiert. Die Bewertung aller generierten Korrespondenzen basiert auf ausgewählten indirekten Metriken. Hierzu gehört auch die Performance bei konkreten Anwendungsfällen wie der Lokalisierung von Landmarken und dem Erstellen von statistischen Formmodellen.
Im Rahmen der Arbeit wird gezeigt, dass FlowSSM Korrespondenzen produziert, deren Qualität dem aktuellen State-of-the-art entspricht. Durch das Hinzufügen der zweiten Kostenfunktion wird die Qualität der Korrespondenzen auf einem globalen Level noch weiter gesteigert. Prinzipiell lässt sich jedoch keine Hierarchie zwischen den Methoden ableiten, da die Performance stark innerhalb der untersuchten Metriken und Datensätzen schwankt. Die Auswahl einer passenden Methode sollte sich daher vor allem am Anwendungsfall orientieren
Construction of a Test Library for the Rolling Stock Rotation Problem with Predictive Maintenance
No full textWe describe the development of a test library for the rolling stock rotation problem with predictive maintenance (RSRP-PdM). Our approach involves the utilization of genuine timetables from a private German railroad company. The generated instances incorporate probability distribution functions for modeling the health states of the vehicles and the considered trips possess varying degradation functions. RSRP-PdM involves assigning trips to a fleet of vehicles and scheduling their maintenance based on their individual health states. The goal is to minimize the total costs consisting of operational costs and the expected costs associated with vehicle failures. The failure probability is dependent on the health states of the vehicles, which are assumed to be random variables distributed by a family of probability distributions. Each distribution is represented by the parameters characterizing it and during the operation of the trips, these parameters get altered. Our approach incorporates non-linear degradation functions to describe the inference of the parameters but also linear ones could be applied. The resulting instances consist of the timetables of the individual lines that use the same vehicle type. Overall, we employ these assumptions and utilize open-source data to create a library of instances with varying difficulty. Our approach is vital for evaluating and comparing algorithms designed to solve the RSRP-PdM
Automated segmentation of insect anatomy from micro-CT images using deep learning
No full textThree-dimensional (3D) imaging, such as micro-computed tomography (micro-CT), is increasingly being used by organismal biologists for precise and comprehensive anatomical characterization. However, the segmentation of anatomical structures remains a bottleneck in research, often requiring tedious manual work. Here, we propose a pipeline for the fully-automated segmentation of anatomical structures in micro-CT images utilizing state-of-the-art deep learning methods, selecting the ant brain as a test case. We implemented the U-Net architecture for 2D image segmentation for our convolutional neural network (CNN), combined with pixel-island detection. For training and validation of the network, we assembled a dataset of semi-manually segmented brain images of 76 ant species. The trained network predicted the brain area in ant images fast and accurately; its performance tested on validation sets showed good agreement between the prediction and the target, scoring 80% Intersection over Union
(IoU) and 90% Dice Coefficient (F1) accuracy. While manual segmentation usually takes many hours for each brain, the trained network takes only a few minutes. Furthermore, our network is generalizable for segmenting the whole neural system in full-body scans, and works in tests on distantly related and morphologically divergent insects (e.g., fruit flies). The latter suggests that methods like the one presented here generally apply across diverse taxa. Our method makes the construction of segmented maps and the morphological quantification of different species more efficient and scalable to large datasets, a step toward a big data approach to organismal anatomy
