@phdthesis{Yu2024,
  author    = {Yu, Yanying},
  title     = {Applied machine learning for the analysis of CRISPR-Cas systems},
  doi       = {10.25972/OPUS-32021},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-320219},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Among the defense strategies developed in microbes over millions of years, the innate adaptive CRISPR-Cas immune systems have spread across most of bacteria and archaea. The flexibility, simplicity, and specificity of CRISPR-Cas systems have laid the foundation for CRISPR-based genetic tools. Yet, the efficient administration of CRISPR-based tools demands rational designs to maximize the on-target efficiency and off-target specificity. Specifically, the selection of guide RNAs (gRNAs), which play a crucial role in the target recognition of CRISPR-Cas systems, is non-trivial. Despite the fact that the emerging machine learning techniques provide a solution to aid in gRNA design with prediction algorithms, design rules for many CRISPR-Cas systems are ill-defined, hindering their broader applications. CRISPR interference (CRISPRi), an alternative gene silencing technique using a catalytically dead Cas protein to interfere with transcription, is a leading technique in bacteria for functional interrogation, pathway manipulation, and genome-wide screens. Although the application is promising, it also is hindered by under-investigated design rules. Therefore, in this work, I develop a state-of-art predictive machine learning model for guide silencing efficiency in bacteria leveraging the advantages of feature engineering, data integration, interpretable AI, and automated machine learning. I first systematically investigate the influential factors that attribute to the extent of depletion in multiple CRISPRi genome-wide essentiality screens in Escherichia coli and demonstrate the surprising dominant contribution of gene-specific effects, such as gene expression level. These observations allowed me to segregate the confounding gene-specific effects using a mixed-effect random forest (MERF) model to provide a better estimate of guide efficiency, together with the improvement led by integrating multiple screens. The MERF model outperformed existing tools in an independent high-throughput saturating screen. I next interpret the predictive model to extract the design rules for robust gene silencing, such as the preference for cytosine and disfavoring for guanine and thymine within and around the protospacer adjacent motif (PAM) sequence. I further incorporated the MERF model in a web-based tool that is freely accessible at www.ciao.helmholtz-hiri.de. When comparing the MERF model with existing tools, the performance of the alternative gRNA design tool optimized for CRISPRi in eukaryotes when applied to bacteria was far from satisfying, questioning the robustness of prediction algorithms across organisms. In addition, the CRISPR-Cas systems exhibit diverse mechanisms albeit with some similarities. The captured predictive patterns from one dataset thereby are at risk of poor generalization when applied across organisms and CRISPR-Cas techniques. To fill the gap, the machine learning approach I present here for CRISPRi could serve as a blueprint for the effective development of prediction algorithms for specific organisms or CRISPR-Cas systems of interest. The explicit workflow includes three principle steps: 1) accommodating the feature set for the CRISPR-Cas system or technique; 2) optimizing a machine learning model using automated machine learning; 3) explaining the model using interpretable AI. To illustrate the applicability of the workflow and diversity of results when applied across different bacteria and CRISPR-Cas systems, I have applied this workflow to analyze three distinct CRISPR-Cas genome-wide screens. From the CRISPR base editor essentiality screen in E. coli, I have determined the PAM preference and sequence context in the editing window for efficient editing, such as A at the 2nd position of PAM, A/TT/TG downstream of PAM, and TC at the 4th to 5th position of gRNAs. From the CRISPR-Cas13a screen in E. coli, in addition to the strong correlation with the guide depletion, the target expression level is the strongest predictor in the model, supporting it as a main determinant of the activation of Cas13-induced immunity and better characterizing the CRISPR-Cas13 system. From the CRISPR-Cas12a screen in Klebsiella pneumoniae, I have extracted the design rules for robust antimicrobial activity across K. pneumoniae strains and provided a predictive algorithm for gRNA design, facilitating CRISPR-Cas12a as an alternative technique to tackle antibiotic resistance. Overall, this thesis presents an accurate prediction algorithm for CRISPRi guide efficiency in bacteria, providing insights into the determinants of efficient silencing and guide designs. The systematic exploration has led to a robust machine learning approach for effective model development in other bacteria and CRISPR-Cas systems. Applying the approach in the analysis of independent CRISPR-Cas screens not only sheds light on the design rules but also the mechanisms of the CRISPR-Cas systems. Together, I demonstrate that applied machine learning paves the way to a deeper understanding and a broader application of CRISPR-Cas systems.},
  subject      = {Maschinelles Lernen},
  language  = {en}
}
@phdthesis{deGraafgebButtler2024,
  author    = {de Graaf [geb. Buttler], Simone Linda},
  title     = {From Small to Large Data: Leveraging Synthetic Data for Inventory Management},
  doi       = {10.25972/OPUS-36136},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-361364},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {In a world of constant change, uncertainty has become a daily challenge for businesses. Rapidly shifting market conditions highlight the need for flexible responses to unforeseen events. Operations Management (OM) is crucial for optimizing business processes, including site planning, production control, and inventory management. Traditionally, companies have relied on theoretical models from microeconomics, game theory, optimization, and simulation. However, advancements in machine learning and mathematical optimization have led to a new research field: data-driven OM. Data-driven OM uses real data, especially time series data, to create more realistic models that better capture decision-making complexities. Despite the promise of this new research area, a significant challenge remains: the availability of extensive historical training data. Synthetic data, which mimics real data, has been used to address this issue in other machine learning applications. Therefore, this dissertation explores how synthetic data can be leveraged to improve decisions for data-driven inventory management, focusing on the single-period newsvendor problem, a classic stochastic optimization problem in inventory management. The first article, "A Meta Analysis of Data-Driven Newsvendor Approaches", presents a standardized evaluation framework for data-driven prescriptive approaches, tested through a numerical study. Findings suggest model performance is not robust, emphasizing the need for a standardized evaluation process. The second article, "Application of Generative Adversarial Networks in Inventory Management", examines using synthetic data generated by Generative Adversarial Networks (GANs) for the newsvendor problem. This study shows GANs can model complex demand relationships, offering a promising alternative to traditional methods. The third article, "Combining Synthetic Data and Transfer Learning for Deep Reinforcement Learning in Inventory Management", proposes a method using Deep Reinforcement Learning (DRL) with synthetic and real data through transfer learning. This approach trains a generative model to learn demand distributions, generates synthetic data, and fine-tunes a DRL agent on a smaller real dataset. This method outperforms traditional approaches in controlled and practical settings, though further research is needed to generalize these findings.},
  subject      = {Bestandsmanagement},
  language  = {en}
}
@phdthesis{Allgaier2024,
  author    = {Allgaier, Johannes},
  title     = {Machine Learning Explainability on Multi-Modal Data using Ecological Momentary Assessments in the Medical Domain},
  doi       = {10.25972/OPUS-35118},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-351189},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Introduction. Mobile health (mHealth) integrates mobile devices into healthcare, enabling remote monitoring, data collection, and personalized interventions. Machine Learning (ML), a subfield of Artificial Intelligence (AI), can use mHealth data to confirm or extend domain knowledge by finding associations within the data, i.e., with the goal of improving healthcare decisions. In this work, two data collection techniques were used for mHealth data fed into ML systems: Mobile Crowdsensing (MCS), which is a collaborative data gathering approach, and Ecological Momentary Assessments (EMA), which capture real-time individual experiences within the individual's common environments using questionnaires and sensors. We collected EMA and MCS data on tinnitus and COVID-19. About 15 \% of the world's population suffers from tinnitus. Materials \& Methods. This thesis investigates the challenges of ML systems when using MCS and EMA data. It asks: How can ML confirm or broad domain knowledge? Domain knowledge refers to expertise and understanding in a specific field, gained through experience and education. Are ML systems always superior to simple heuristics and if yes, how can one reach explainable AI (XAI) in the presence of mHealth data? An XAI method enables a human to understand why a model makes certain predictions. Finally, which guidelines can be beneficial for the use of ML within the mHealth domain? In tinnitus research, ML discerns gender, temperature, and season-related variations among patients. In the realm of COVID-19, we collaboratively designed a COVID-19 check app for public education, incorporating EMA data to offer informative feedback on COVID-19-related matters. This thesis uses seven EMA datasets with more than 250,000 assessments. Our analyses revealed a set of challenges: App user over-representation, time gaps, identity ambiguity, and operating system specific rounding errors, among others. Our systematic review of 450 medical studies assessed prior utilization of XAI methods. Results. ML models predict gender and tinnitus perception, validating gender-linked tinnitus disparities. Using season and temperature to predict tinnitus shows the association of these variables with tinnitus. Multiple assessments of one app user can constitute a group. Neglecting these groups in data sets leads to model overfitting. In select instances, heuristics outperform ML models, highlighting the need for domain expert consultation to unveil hidden groups or find simple heuristics. Conclusion. This thesis suggests guidelines for mHealth related data analyses and improves estimates for ML performance. Close communication with medical domain experts to identify latent user subsets and incremental benefits of ML is essential.},
  subject      = {Maschinelles Lernen},
  language  = {en}
}
@phdthesis{Kobs2024,
  author    = {Kobs, Konstantin},
  title     = {Think outside the Black Box: Model-Agnostic Deep Learning with Domain Knowledge},
  doi       = {10.25972/OPUS-34968},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-349689},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Deep Learning (DL) models are trained on a downstream task by feeding (potentially preprocessed) input data through a trainable Neural Network (NN) and updating its parameters to minimize the loss function between the predicted and the desired output. While this general framework has mainly remained unchanged over the years, the architectures of the trainable models have greatly evolved. Even though it is undoubtedly important to choose the right architecture, we argue that it is also beneficial to develop methods that address other components of the training process. We hypothesize that utilizing domain knowledge can be helpful to improve DL models in terms of performance and/or efficiency. Such model-agnostic methods can be applied to any existing or future architecture. Furthermore, the black box nature of DL models motivates the development of techniques to understand their inner workings. Considering the rapid advancement of DL architectures, it is again crucial to develop model-agnostic methods. In this thesis, we explore six principles that incorporate domain knowledge to understand or improve models. They are applied either on the input or output side of the trainable model. Each principle is applied to at least two DL tasks, leading to task-specific implementations. To understand DL models, we propose to use Generated Input Data coming from a controllable generation process requiring knowledge about the data properties. This way, we can understand the model's behavior by analyzing how it changes when one specific high-level input feature changes in the generated data. On the output side, Gradient-Based Attribution methods create a gradient at the end of the NN and then propagate it back to the input, indicating which low-level input features have a large influence on the model's prediction. The resulting input features can be interpreted by humans using domain knowledge. To improve the trainable model in terms of downstream performance, data and compute efficiency, or robustness to unwanted features, we explore principles that each address one of the training components besides the trainable model. Input Masking and Augmentation directly modifies the training input data, integrating knowledge about the data and its impact on the model's output. We also explore the use of Feature Extraction using Pretrained Multimodal Models which can be seen as a beneficial preprocessing step to extract useful features. When no training data is available for the downstream task, using such features and domain knowledge expressed in other modalities can result in a Zero-Shot Learning (ZSL) setting, completely eliminating the trainable model. The Weak Label Generation principle produces new desired outputs using knowledge about the labels, giving either a good pretraining or even exclusive training dataset to solve the downstream task. Finally, improving and choosing the right Loss Function is another principle we explore in this thesis. Here, we enrich existing loss functions with knowledge about label interactions or utilize and combine multiple task-specific loss functions in a multitask setting. We apply the principles to classification, regression, and representation tasks as well as to image and text modalities. We propose, apply, and evaluate existing and novel methods to understand and improve the model. Overall, this thesis introduces and evaluates methods that complement the development and choice of DL model architectures.},
  subject      = {Deep learning},
  language  = {en}
}
@phdthesis{Nadernezhad2024,
  author    = {Nadernezhad, Ali},
  title     = {Engineering approaches in biofabrication of vascularized structures},
  doi       = {10.25972/OPUS-34589},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-345892},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Biofabrication technologies must address numerous parameters and conditions to reconstruct tissue complexity in vitro. A critical challenge is vascularization, especially for large constructs exceeding diffusion limits. This requires the creation of artificial vascular structures, a task demanding the convergence and integration of multiple engineering approaches. This doctoral dissertation aims to achieve two primary objectives: firstly, to implement and refine engineering methods for creating artificial microvascular structures using Melt Electrowriting (MEW)-assisted sacrificial templating, and secondly, to deepen the understanding of the critical factors influencing the printability of bioink formulations in 3D extrusion bioprinting. In the first part of this dissertation, two innovative sacrificial templating techniques using MEW are explored. Utilizing a carbohydrate glass as a fugitive material, a pioneering advancement in the processing of sugars with MEW with a resolution under 100 microns was made. Furthermore, by introducing the "print-and-fuse" strategy as a groundbreaking method, biomimetic branching microchannels embedded in hydrogel matrices were fabricated, which can then be endothelialized to mirror in vivo vascular conditions. The second part of the dissertation explores extrusion bioprinting. By introducing a simple binary bioink formulation, the correlation between physical properties and printability was showcased. In the next step, employing state-of-the-art machine-learning approaches revealed a deeper understanding of the correlations between bioink properties and printability in an extended library of hydrogel formulations. This dissertation offers in-depth insights into two key biofabrication technologies. Future work could merge these into hybrid methods for the fabrication of vascularized constructs, combining MEW's precision with fine-tuned bioink properties in automated extrusion bioprinting.},
  subject      = {3D-Druck},
  language  = {en}
}
@phdthesis{Kleineisel2024,
  author    = {Kleineisel, Jonas},
  title     = {Variational networks in magnetic resonance imaging - Application to spiral cardiac MRI and investigations on image quality},
  doi       = {10.25972/OPUS-34737},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-347370},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2024},
  abstract  = {Acceleration is a central aim of clinical and technical research in magnetic resonance imaging (MRI) today, with the potential to increase robustness, accessibility and patient comfort, reduce cost, and enable entirely new kinds of examinations. A key component in this endeavor is image reconstruction, as most modern approaches build on advanced signal and image processing. Here, deep learning (DL)-based methods have recently shown considerable potential, with numerous publications demonstrating benefits for MRI reconstruction. However, these methods often come at the cost of an increased risk for subtle yet critical errors. Therefore, the aim of this thesis is to advance DL-based MRI reconstruction, while ensuring high quality and fidelity with measured data. A network architecture specifically suited for this purpose is the variational network (VN). To investigate the benefits these can bring to non-Cartesian cardiac imaging, the first part presents an application of VNs, which were specifically adapted to the reconstruction of accelerated spiral acquisitions. The proposed method is compared to a segmented exam, a U-Net and a compressed sensing (CS) model using qualitative and quantitative measures. While the U-Net performed poorly, the VN as well as the CS reconstruction showed good output quality. In functional cardiac imaging, the proposed real-time method with VN reconstruction substantially accelerates examinations over the gold-standard, from over 10 to just 1 minute. Clinical parameters agreed on average. Generally in MRI reconstruction, the assessment of image quality is complex, in particular for modern non-linear methods. Therefore, advanced techniques for precise evaluation of quality were subsequently demonstrated. With two distinct methods, resolution and amplification or suppression of noise are quantified locally in each pixel of a reconstruction. Using these, local maps of resolution and noise in parallel imaging (GRAPPA), CS, U-Net and VN reconstructions were determined for MR images of the brain. In the tested images, GRAPPA delivers uniform and ideal resolution, but amplifies noise noticeably. The other methods adapt their behavior to image structure, where different levels of local blurring were observed at edges compared to homogeneous areas, and noise was suppressed except at edges. Overall, VNs were found to combine a number of advantageous properties, including a good trade-off between resolution and noise, fast reconstruction times, and high overall image quality and fidelity of the produced output. Therefore, this network architecture seems highly promising for MRI reconstruction.},
  subject      = {Kernspintomografie},
  language  = {en}
}
@phdthesis{Marquardt2023,
  author    = {Marquardt, Andr{\´e}},
  title     = {Machine-Learning-Based Identification of Tumor Entities, Tumor Subgroups, and Therapy Options},
  doi       = {10.25972/OPUS-32954},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-329548},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Molecular genetic analyses, such as mutation analyses, are becoming increasingly important in the tumor field, especially in the context of therapy stratification. The identification of the underlying tumor entity is crucial, but can sometimes be difficult, for example in the case of metastases or the so-called Cancer of Unknown Primary (CUP) syndrome. In recent years, methylome and transcriptome utilizing machine learning (ML) approaches have been developed to enable fast and reliable tumor and tumor subtype identification. However, so far only methylome analysis have become widely used in routine diagnostics. The present work addresses the utility of publicly available RNA-sequencing data to determine the underlying tumor entity, possible subgroups, and potential therapy options. Identification of these by ML - in particular random forest (RF) models - was the first task. The results with test accuracies of up to 99\% provided new, previously unknown insights into the trained models and the corresponding entity prediction. Reducing the input data to the top 100 mRNA transcripts resulted in a minimal loss of prediction quality and could potentially enable application in clinical or real-world settings. By introducing the ratios of these top 100 genes to each other as a new database for RF models, a novel method was developed enabling the use of trained RF models on data from other sources. Further analysis of the transcriptomic differences of metastatic samples by visual clustering showed that there were no differences specific for the site of metastasis. Similarly, no distinct clusters were detectable when investigating primary tumors and metastases of cutaneous skin melanoma (SKCM). Subsequently, more than half of the validation datasets had a prediction accuracy of at least 80\%, with many datasets even achieving a prediction accuracy of - or close to - 100\%. To investigate the applicability of the used methods for subgroup identification, the TCGA-KIPAN dataset, consisting of the three major kidney cancer subgroups, was used. The results revealed a new, previously unknown subgroup consisting of all histopathological groups with clinically relevant characteristics, such as significantly different survival. Based on significant differences in gene expression, potential therapeutic options of the identified subgroup could be proposed. Concludingly, in exploring the potential applicability of RNA-sequencing data as a basis for therapy prediction, it was shown that this type of data is suitable to predict entities as well as subgroups with high accuracy. Clinical relevance was also demonstrated for a novel subgroup in renal cell carcinoma. The reduction of the number of genes required for entity prediction to 100 genes, enables panel sequencing and thus demonstrates potential applicability in a real-life setting.},
  subject      = {Maschinelles Lernen},
  language  = {en}
}
@phdthesis{Krenzer2023,
  author    = {Krenzer, Adrian},
  title     = {Machine learning to support physicians in endoscopic examinations with a focus on automatic polyp detection in images and videos},
  doi       = {10.25972/OPUS-31911},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-319119},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Deep learning enables enormous progress in many computer vision-related tasks. Artificial Intel- ligence (AI) steadily yields new state-of-the-art results in the field of detection and classification. Thereby AI performance equals or exceeds human performance. Those achievements impacted many domains, including medical applications. One particular field of medical applications is gastroenterology. In gastroenterology, machine learning algorithms are used to assist examiners during interventions. One of the most critical concerns for gastroenterologists is the development of Colorectal Cancer (CRC), which is one of the leading causes of cancer-related deaths worldwide. Detecting polyps in screening colonoscopies is the essential procedure to prevent CRC. Thereby, the gastroenterologist uses an endoscope to screen the whole colon to find polyps during a colonoscopy. Polyps are mucosal growths that can vary in severity. This thesis supports gastroenterologists in their examinations with automated detection and clas- sification systems for polyps. The main contribution is a real-time polyp detection system. This system is ready to be installed in any gastroenterology practice worldwide using open-source soft- ware. The system achieves state-of-the-art detection results and is currently evaluated in a clinical trial in four different centers in Germany. The thesis presents two additional key contributions: One is a polyp detection system with ex- tended vision tested in an animal trial. Polyps often hide behind folds or in uninvestigated areas. Therefore, the polyp detection system with extended vision uses an endoscope assisted by two additional cameras to see behind those folds. If a polyp is detected, the endoscopist receives a vi- sual signal. While the detection system handles the additional two camera inputs, the endoscopist focuses on the main camera as usual. The second one are two polyp classification models, one for the classification based on shape (Paris) and the other on surface and texture (NBI International Colorectal Endoscopic (NICE) classification). Both classifications help the endoscopist with the treatment of and the decisions about the detected polyp. The key algorithms of the thesis achieve state-of-the-art performance. Outstandingly, the polyp detection system tested on a highly demanding video data set shows an F1 score of 90.25 \% while working in real-time. The results exceed all real-time systems in the literature. Furthermore, the first preliminary results of the clinical trial of the polyp detection system suggest a high Adenoma Detection Rate (ADR). In the preliminary study, all polyps were detected by the polyp detection system, and the system achieved a high usability score of 96.3 (max 100). The Paris classification model achieved an F1 score of 89.35 \% which is state-of-the-art. The NICE classification model achieved an F1 score of 81.13 \%. Furthermore, a large data set for polyp detection and classification was created during this thesis. Therefore a fast and robust annotation system called Fast Colonoscopy Annotation Tool (FastCAT) was developed. The system simplifies the annotation process for gastroenterologists. Thereby the i gastroenterologists only annotate key parts of the endoscopic video. Afterward, those video parts are pre-labeled by a polyp detection AI to speed up the process. After the AI has pre-labeled the frames, non-experts correct and finish the annotation. This annotation process is fast and ensures high quality. FastCAT reduces the overall workload of the gastroenterologist on average by a factor of 20 compared to an open-source state-of-art annotation tool.},
  subject      = {Deep Learning},
  language  = {en}
}
@phdthesis{Steininger2023,
  author    = {Steininger, Michael},
  title     = {Deep Learning for Geospatial Environmental Regression},
  doi       = {10.25972/OPUS-31312},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-313121},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {Environmental issues have emerged especially since humans burned fossil fuels, which led to air pollution and climate change that harm the environment. These issues' substantial consequences evoked strong efforts towards assessing the state of our environment. Various environmental machine learning (ML) tasks aid these efforts. These tasks concern environmental data but are common ML tasks otherwise, i.e., datasets are split (training, validatition, test), hyperparameters are optimized on validation data, and test set metrics measure a model's generalizability. This work focuses on the following environmental ML tasks: Regarding air pollution, land use regression (LUR) estimates air pollutant concentrations at locations where no measurements are available based on measured locations and each location's land use (e.g., industry, streets). For LUR, this work uses data from London (modeled) and Zurich (measured). Concerning climate change, a common ML task is model output statistics (MOS), where a climate model's output for a study area is altered to better fit Earth observations and provide more accurate climate data. This work uses the regional climate model (RCM) REMO and Earth observations from the E-OBS dataset for MOS. Another task regarding climate is grain size distribution interpolation where soil properties at locations without measurements are estimated based on the few measured locations. This can provide climate models with soil information, that is important for hydrology. For this task, data from Lower Franconia is used. Such environmental ML tasks commonly have a number of properties: (i) geospatiality, i.e., their data refers to locations relative to the Earth's surface. (ii) The environmental variables to estimate or predict are usually continuous. (iii) Data can be imbalanced due to relatively rare extreme events (e.g., extreme precipitation). (iv) Multiple related potential target variables can be available per location, since measurement devices often contain different sensors. (v) Labels are spatially often only sparsely available since conducting measurements at all locations of interest is usually infeasible. These properties present challenges but also opportunities when designing ML methods for such tasks. In the past, environmental ML tasks have been tackled with conventional ML methods, such as linear regression or random forests (RFs). However, the field of ML has made tremendous leaps beyond these classic models through deep learning (DL). In DL, models use multiple layers of neurons, producing increasingly higher-level feature representations with growing layer depth. DL has made previously infeasible ML tasks feasible, improved the performance for many tasks in comparison to existing ML models significantly, and eliminated the need for manual feature engineering in some domains due to its ability to learn features from raw data. To harness these advantages for environmental domains it is promising to develop novel DL methods for environmental ML tasks. This thesis presents methods for dealing with special challenges and exploiting opportunities inherent to environmental ML tasks in conjunction with DL. To this end, the proposed methods explore the following techniques: (i) Convolutions as in convolutional neural networks (CNNs) to exploit reoccurring spatial patterns in geospatial data. (ii) Posing the problems as regression tasks to estimate the continuous variables. (iii) Density-based weighting to improve estimation performance for rare and extreme events. (iv) Multi-task learning to make use of multiple related target variables. (v) Semi-supervised learning to cope with label sparsity. Using these techniques, this thesis considers four research questions: (i) Can air pollution be estimated without manual feature engineering? This is answered positively by the introduction of the CNN-based LUR model MapLUR as well as the off-the-shelf LUR solution OpenLUR. (ii) Can colocated pollution data improve spatial air pollution models? Multi-task learning for LUR is developed for this, showing potential for improvements with colocated data. (iii) Can DL models improve the quality of climate model outputs? The proposed DL climate MOS architecture ConvMOS demonstrates this. Additionally, semi-supervised training of multilayer perceptrons (MLPs) for grain size distribution interpolation is presented, which can provide improved input data. (iv) Can DL models be taught to better estimate climate extremes? To this end, density-based weighting for imbalanced regression (DenseLoss) is proposed and applied to the DL architecture ConvMOS, improving climate extremes estimation. These methods show how especially DL techniques can be developed for environmental ML tasks with their special characteristics in mind. This allows for better models than previously possible with conventional ML, leading to more accurate assessment and better understanding of the state of our environment.},
  subject      = {Deep learning},
  language  = {en}
}
@phdthesis{Gold2023,
  author    = {Gold, Lukas},
  title     = {Methods for the state estimation of lithium-ion batteries},
  doi       = {10.25972/OPUS-30618},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-306180},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2023},
  abstract  = {This work introduced the reader to all relevant fields to tap into an ultrasound-based state of charge estimation and provides a blueprint for the procedure to achieve and test the fundamentals of such an approach. It spanned from an in-depth electrochemical characterization of the studied battery cells over establishing the measurement technique, digital processing of ultrasonic transmission signals, and characterization of the SoC dependent property changes of those signals to a proof of concept of an ultrasound-based state of charge estimation. The State of the art \& theoretical background chapter focused on the battery section on the mechanical property changes of lithium-ion batteries during operation. The components and the processes involved to manufacture a battery cell were described to establish the fundamentals for later interrogation. A comprehensive summary of methods for state estimation was given and an emphasis was laid on mechanical methods, including a critical review of the most recent research on ultrasound-based state estimation. Afterward, the fundamentals of ultrasonic non-destructive evaluation were introduced, starting with the sound propagation modes in isotropic boundary-free media, followed by the introduction of boundaries and non-isotropic structure to finally approach the class of fluid-saturated porous media, which batteries can be counted to. As the processing of the ultrasonic signals transmitted through lithium-ion battery cells with the aim of feature extraction was one of the main goals of this work, the fundamentals of digital signal processing and methods for the time of flight estimation were reviewed and compared in a separate section. All available information on the interrogated battery cell and the instrumentation was collected in the Experimental methods \& instrumentation chapter, including a detailed step-by-step manual of the process developed in this work to create and attach a sensor stack for ultrasonic interrogation based on low-cost off-the-shelf piezo elements. The Results \& discussion chapter opened with an in-depth electrochemical and post-mortem interrogation to reverse engineer the battery cell design and its internal structure. The combination of inductively coupled plasma-optical emission spectrometry and incremental capacity analysis applied to three-electrode lab cells, constructed from the studied battery cell's materials, allowed to identify the SoC ranges in which phase transitions and staging occur and thereby directly links changes in the ultrasonic signal properties with the state of the active materials, which makes this work stand out among other studies on ultrasound-based state estimation. Additional dilatometer experiments were able to prove that the measured effect in ultrasonic time of flight cannot originate from the thickness increase of the battery cells alone, as this thickness increase is smaller and in opposite direction to the change in time of flight. Therefore, changes in elastic modulus and density have to be responsible for the observed effect. The construction of the sensor stack from off-the-shelf piezo elements, its electromagnetic shielding, and attachment to both sides of the battery cells was treated in a subsequent section. Experiments verified the necessity of shielding and its negligible influence on the ultrasonic signals. A hypothesis describing the metal layer in the pouch foil to be the transport medium of an electrical coupling/distortion between sending and receiving sensor was formulated and tested. Impedance spectroscopy was shown to be a useful tool to characterize the resonant behavior of piezo elements and ensure the mechanical coupling of such to the surface of the battery cells. The excitation of the piezo elements by a raised cosine (RCn) waveform with varied center frequency in the range of 50 kHz to 250 kHz was studied in the frequency domain and the influence of the resonant behavior, as identified prior by impedance spectroscopy, on waveform and frequency content was evaluated to be uncritical. Therefore, the forced oscillation produced by this excitation was assumed to be mechanically coupled as ultrasonic waves into the battery cells. The ultrasonic waves transmitted through the battery cell were recorded by piezo elements on the opposing side. A first inspection of the raw, unprocessed signals identified the transmission of two main wave packages and allowed the identification of two major trends: the time of flight of ultrasonic wave packages decreases with the center frequency of the RCn waveform, and with state of charge. These trends were to be assessed further in the subsequent sections. Therefore, methods for the extraction of features (properties) from the ultrasonic signals were established, compared, and tested in a dedicated section. Several simple and advanced thresholding methods were compared with envelope-based and cross-correlation methods to estimate the time of flight (ToF). It was demonstrated that the envelope-based method yields the most robust estimate for the first and second wave package. This finding is in accordance with the literature stating that an envelope-based method is best suited for dispersive, absorptive media [204], to which lithium-ion batteries are counted. Respective trends were already suggested by the heatmap plots of the raw signals vs. RCn frequency and SoC. To enable such a robust estimate, an FIR filter had to be designed to preprocess the transmitted signals and thereby attenuate frequency components that verifiably lead to a distorted shape of the envelope. With a robust ToF estimation method selected, the characterization of the signal properties ToF and transmitted energy content (EC) was performed in-depth. A study of cycle-to-cycle variations unveiled that the signal properties are affected by a long rest period and the associated relaxation of the multi-particle system "battery cell" to equilibrium. In detail, during cycling, the signal properties don't reach the same value at a given SoC in two subsequent cycles if the first of the two cycles follows a long rest period. In accordance with the literature, a break-in period, making up for more than ten cycles post-formation, was observed. During this break-in period, the mechanical properties of the system are said to change until a steady state is reached [25]. Experiments at different C-rate showed that ultrasonic signal properties can sense the non-equilibrium state of a battery cell, characterized by an increasing area between charge and discharge curve of the respective signal property vs. SoC plot. This non-equilibrium state relaxes in the rest period following the discharge after the cut-off voltage is reached. The relaxation in the rest period following the charge is much smaller and shows little C-rate dependency as the state is prepared by constant voltage charging at the end of charge voltage. For a purely statistical SoC estimation approach, as employed in this work, where only instantaneous measurements are taken into account and the historic course of the measurement is not utilized as a source of information, the presence of hysteresis and relaxation leads to a reduced estimation accuracy. Future research should address this issue or even utilize the relaxation to improve the estimation accuracy, by incorporating historic information, e.g., by using the derivative of a signal property as an additional feature. The signal properties were then tested for their correlation with SoC as a function of RCn frequency. This allowed identifying trends in the behavior of the signal properties as a function of RCn frequency and C-rate in a condensed fashion and thereby enabled to predict the frequency range, about 50 kHz to 125 kHz, in which the course of the signal properties is best suited for SoC estimation. The final section provided a proof of concept of the ultrasound-based SoC estimation, by applying a support vector regression (SVR) to before thoroughly studied ultrasonic signal properties, as well as current and battery cell voltage. The included case study was split into different parts that assessed the ability of an SVR to estimate the SoC in a variety of scenarios. Seven battery cells, prepared with sensor stacks attached to both faces, were used to generate 14 datasets. First, a comparison of self-tests, where a portion of a dataset is used for training and another for testing, and cross-tests, which use the dataset of one cell for training and the dataset of another for testing, was performed. A root mean square error (RMSE) of 3.9\% to 4.8\% SoC and 3.6\% to 10.0\% SoC was achieved, respectively. In general, it was observed that the SVR is prone to overestimation at low SoCs and underestimation at high SoCs, which was attributed to the pronounced hysteresis and relaxation of the ultrasonic signal properties in this SoC ranges. The fact that higher accuracy is achieved, if the exact cell is known to the model, indicates that a variation between cells exists. This variation between cells can originate from differences in mechanical properties as a result of production variations or from differences in manual sensor placement, mechanical coupling, or resonant behavior of the ultrasonic sensors. To mitigate the effect of the cell-to-cell variations, a test was performed, where the datasets of six out of the seven cells were combined as training data, and the dataset of the seventh cell was used for testing. This reduced the spread of the RMSE from (3.6 - 10.0)\% SoC to (5.9 - 8.5)\% SoC, respectively, once again stating that a databased approach for state estimation becomes more reliable with a large data basis. Utilizing self-tests on seven datasets, the effect of additional features on the state estimation result was tested. The involvement of an additional feature did not necessarily improve the estimation accuracy, but it was shown that a combination of ultrasonic and electrical features is superior to the training with these features alone. To test the ability of the model to estimate the SoC in unknown cycling conditions, a test was performed where the C-rate of the test dataset was not included in the training data. The result suggests that for practical applications it might be sufficient to perform training with the boundary of the use cases in a controlled laboratory environment to handle the estimation in a broad spectrum of use cases. In comparison with literature, this study stands out by utilizing and modifying off-the-shelf piezo elements to equip state-of-the-art lithium-ion battery cells with ultrasonic sensors, employing a range of center frequencies for the waveform, transmitted through the battery cell, instead of a fixed frequency and by allowing the SVR to choose the frequency that yields the best result. The characterization of the ultrasonic signal properties as a function of RCn frequency and SoC and the assignment of characteristic changes in the signal properties to electrochemical processes, such as phase transitions and staging, makes this work unique. By studying a range of use cases, it was demonstrated that an improved SoC estimation accuracy can be achieved with the aid of ultrasonic measurements - thanks to the correlation of the mechanical properties of the battery cells with the SoC.},
  subject      = {Lithium-Ionen-Akkumulator},
  language  = {en}
}