<p>Gaussian Process Regression (GPR) is a widely used surrogate model in efficient global optimization (EGO) due to its capability to provide uncertainty estimates in the prediction. The cost of creating a GPR model for large data sets is high. On the other hand, neural network (NN) models scale better compared to GPR as the number of samples increase. Unfortunately, the uncertainty estimates for NN prediction are not readily available. In this work, a scalable algorithm is developed for EGO using NN-based prediction and uncertainty (EGONN). Initially, two different NNs are created using two different data sets. The first NN models the output based on the input values in the first data set while the second NN models the prediction error of the first NN using the second data set. The next infill point is added to the first data set based on criteria like expected improvement or prediction uncertainty. EGONN is demonstrated on the optimization of the Forrester function and a constrained Branin function and is compared with EGO. The convergence criteria is based on the maximum number of infill points in both cases. The algorithm is able to reach the optimum point within the given budget. The EGONN is extended to handle constraints explicitly and is utilized for aerodynamic shape optimization of the RAE 2822 airfoil in transonic viscous flow at a free-stream Mach number of 0.734 and a Reynolds number of 6.5 million. The results obtained from EGONN are compared with the results from gradient-based optimization (GBO) using adjoints. The optimum shape obtained from EGONN is comparable to the shape obtained from GBO and is able to eliminate the shock. The drag coefficient is reduced from 200 drag counts to 114 and is close to 110 drag counts obtained from GBO. The EGONN is also extended to handle uncertainty quantification (uqEGONN) using prediction uncertainty as an infill method. The convergence criteria is based on the relative change of summary statistics such as mean and standard deviation of an uncertain quantity. The uqEGONN is tested on Ishigami function with an initial sample size of 100 samples and the algorithm terminates after 70 infill points. The statistics obtained from uqEGONN (using only 170 function evaluations) are close to the values obtained from directly evaluating the function one million times. uqEGONN is demonstrated on to quantifying the uncertainty in the airfoil performance due to geometric variations. The algorithm terminates within 100 computational fluid dynamics (CFD) analyses and the statistics obtained from the algorithm are close to the one obtained from 1000 direct CFD based evaluations.</p>

In recent years, advanced machine learning (ML) models have revolutionized industries ranging from the healthcare sector to retail and E-commerce. However, these models have become increasingly complex, making it difficult for even domain experts to understand and retrace the model's decision-making process. To address this challenge, several frameworks for explainable AI have been proposed and developed. This thesis focuses on counterfactual explanations (CFEs), which provide actionable insights by informing users how to modify inputs to achieve desired outputs. However, computing CFEs for a general black-box ML model is computationally expensive since it hinges on solving a challenging optimization problem. To efficiently solve this optimization problem, we propose using Bayesian optimization (BO), and introduce the novel algorithm Separated Bayesian Optimization (SBO). SBO exploits the formulation of the counterfactual function as a composite function. Additionally, we propose warm-starting SBO, which addresses the computational challenges associated with computing multiple CFEs. By decoupling the generation of a surrogate model for the black-box model and the computation of specific CFEs, warm-starting SBO allows us to reuse previous data and computations, resulting in computational discounts and improved efficiency for large-scale applications. Through numerical experiments, we demonstrate that BO is a viable optimization scheme for computing CFEs for black-box ML models. BO achieves computational efficiency while maintaining good accuracy. SBO improves upon this by requiring fewer evaluations while achieving accuracies comparable to the best conventional optimizer tested. Both BO and SBO exhibit improved capabilities in handling various classes of ML decision models compared to the tested baseline optimizers. Finally, Warm-starting SBO significantly enhances the performance of SBO, reducing function evaluations and errors when computing multiple sequential CFEs. The results indicate a strong potential for large-scale industry applications. / Avancerade maskininlärningsmodeller (ML-modeller) har på senaste åren haft stora framgångar inom flera delar av näringslivet, med allt ifrån hälso- och sjukvårdssektorn till detaljhandel och e-handel. I jämn takt med denna utveckling har det dock även kommit en ökad komplexitet av dessa ML-modeller vilket nu lett till att även domänexperter har svårigheter med att förstå och tolka modellernas beslutsprocesser. För att bemöta detta problem har flertalet förklarbar AI ramverk utvecklats. Denna avhandling fokuserar på kontrafaktuella förklaringar (CFEs). Detta är en förklaringstyp som anger för användaren hur denne bör modifiera sin indata för att uppnå ett visst modellbeslut. För en generell svarta-låda ML-modell är dock beräkningsmässigt kostsamt att beräkna CFEs då det krävs att man löser ett utmanande optimeringsproblem. För att lösa optimeringsproblemet föreslår vi användningen av Bayesiansk Optimering (BO), samt presenterar den nya algoritmen Separated Bayesian Optimization (SBO). SBO utnyttjar kompositionsformuleringen av den kontrafaktuella funktionen. Vidare, utforskar vi beräkningen av flera sekventiella CFEs för vilket vi presenterar varm-startad SBO. Varm-startad SBO lyckas återanvända data samt beräkningar från tidigare CFEs tack vare en separation av surrogat-modellen för svarta-låda ML-modellen och beräkningen av enskilda CFEs. Denna egenskap leder till en minskad beräkningskostnad samt ökad effektivitet för storskaliga tillämpningar.  I de genomförda experimenten visar vi att BO är en lämplig optimeringsmetod för att beräkna CFEs för svarta-låda ML-modeller tack vare en god beräknings effektivitet kombinerat med hög noggrannhet. SBO presterade ännu bättre med i snitt färre funktionsutvärderingar och med fel nivåer jämförbara med den bästa testade konventionella optimeringsmetoden. Både BO och SBO visade på bättre kapacitet att hantera olika klasser av ML-modeller än de andra testade metoderna. Slutligen observerade vi att varm-startad SBO gav ytterligare prestandaökningar med både minskade funktionsutvärderingar och fel när flera CFEs beräknades. Dessa resultat pekar på stor potential för storskaliga tillämpningar inom näringslivet.

Apache Spark is a popular open-source distributed processing framework that enables efficient processing of large amounts of data. Apache Spark has a large number of configuration parameters that are strongly related to performance. Selecting an optimal configuration for Apache Spark application deployed in a cloud environment is a complex task. Making a poor choice may not only result in poor performance but also increases costs. Manually adjusting the Apache Spark configuration parameters can take a lot of time and may not lead to the best outcomes, particularly in a cloud environment where computing resources are allocated dynamically, and workloads can fluctuate significantly. The focus of this thesis project is the development of an auto-tuning approach for Apache Spark configuration parameters. Four machine learning models are formulated and evaluated to predict Apache Spark’s performance. Additionally, two models for Apache Spark configuration parameter search are created and evaluated to identify the most suitable parameters, resulting in the shortest execution time. The obtained results demonstrates that with the developed auto-tuning approach and adjusting Apache Spark configuration parameters, Apache Spark applications can achieve a shorter execution time than when using the default parameters. The developed auto-tuning approach gives an improved cluster utilization and shorter job execution time, with an average performance improvement of 49.98%, 53.84%, and 64.16% for the three different types of Apache Spark applications benchmarked. / Apache Spark är en populär öppen källkodslösning för distribuerad databehandling som möjliggör effektiv bearbetning av stora mängder data. Apache Spark har ett stort antal konfigurationsparametrar som starkt påverkar prestandan. Att välja en optimal konfiguration för en Apache Spark-applikation som distribueras i en molnmiljö är en komplex uppgift. Ett dåligt val kan inte bara leda till dålig prestanda utan också ökade kostnader. Manuell anpassning av Apache Spark-konfigurationsparametrar kan ta mycket tid och leda till suboptimala resultat, särskilt i en molnmiljö där beräkningsresurser tilldelas dynamiskt och arbetsbelastningen kan variera avsevärt. Fokus för detta examensprojekt är att utveckla en automatisk optimeringsmetod för konfigurationsparametrarna i Apache Spark. Fyra maskininlärningsmodeller formuleras och utvärderas för att förutsäga Apache Sparks prestanda. Dessutom skapas och utvärderas två modeller för att söka efter de mest lämpliga konfigurationsparametrarna för Apache Spark, vilket resulterar i kortast möjliga exekveringstid. De erhållna resultaten visar att den utvecklade automatiska optimeringsmetoden, med anpassning av Apache Sparks konfigurationsparameterar, bidrar till att Apache Spark-applikationer kan uppnå kortare exekveringstider än vid användning av standard-parametrar. Den utvecklade metoden för automatisk optimering bidrar till en förbättrad användning av klustret och kortare exekveringstider, med en genomsnittlig prestandaförbättring på 49,98%, 53,84% och 64,16% för de tre olika typerna av Apache Spark-applikationer som testades.

<p>Co and Ni are two major elements in high temperature structural alloys that include superalloys for turbine engines and hard metals for cutting tools. The recent development of complex concentrated alloys (CCAs), loosely defined as alloys without a single principal element (e.g. CoNiFeMn), offers additional opportunities in designing new alloys through extensive composition and structure modifications. Within CCAs and Co- and Ni-based superalloys, solid-solution strengthening and stacking fault energy engineering are two of the most important strengthening mechanisms. While studied for decades, the potency and quantitative materials properties of these mechanisms remain elusive. </p> <p><br></p> <p>Solid-solution strengthening originates from stress field interactions between dislocations and solute of various species in the alloy. These stress fields can be engineered by composition modification in CCAs, and therefore a wide range of alloys with promising mechanical strength may be designed. This thesis initially reports on experimental and computational validation of newly developed theories for solid-solution strengthening in 3d transition metal (MnFeCoNi) alloys. The strengthening effects of Al, Ti, V, Cr, Cu and Mo as alloying elements are quantified by coupling the Labusch-type strengthening model and experimental measurements. With large atomic misfits with the base alloy, Al, Ti, Mo, and Cr present strong strengthening effects comparable to other Cantor alloys. </p> <p> </p> <p>Stacking fault energy engineering can enable novel deformation mechanisms and exceptional strength in face-centered cubic (FCC) materials such as austenitic TRIP/TWIP steels and CoNi-based superalloys exhibiting local phase transformation strengthening via Suzuki segregation. We employed first-principles calculations to investigate the Suzuki segregation and stacking fault energy of the FCC Co-Ni binary alloys at finite temperatures and concentrations. We quantitatively predicted the Co segregation in the innermost plane of the intrinsic stacking fault (ISF). We further quantified the decrease of stacking fault energy due to segregation.  </p> <p><br></p> <p>We further investigated the driving force of segregation and the origin of the segregation behaviors of 3d, 4d and 5d elements in the Co- and Ni-alloys. Using first-principles calculations, we calculated the ground-state solute-ISF interaction energies and revealed the trends across the periodic table. We discussed the relationships between the interaction energies and the local lattice distortions, charge density redistribution, density of states and local magnetization of the solutes. </p> <p><br></p> <p>Finally, this thesis reports on new methodologies to accelerate first-principles calculations utilizing active learning techniques, such as Bayesian optimization, to efficiently search for the ground-state energy line of the system with limited computational resources. Based on the expected improvement method, new acquisition strategies were developed and will be compared and presented. </p>

[ES] El acceso a una atención sanitaria de alta calidad es un marcador importante del desarrollo de las sociedades humanas. Los aportes tecnológicos a la medicina han mostrado un potencial relevante para descubrir procedimientos efectivos a nivel preventivo, diagnóstico y terapéutico. En particular, los métodos computacionales permiten el procesamiento eficaz de datos médicos y, por tanto, pueden modelar sistemas biológicos complejos. Esto ha influido en el desarrollo de la Medicina Personalizada (MP) durante las últimas décadas, donde la obtención de conocimiento específico de cada caso permite realizar intervenciones a medida, todo ello a un coste de recursos accesible. La simulación de remodelación ósea es un campo prometedor en el contexto de la MP. Predecir un proceso de adaptación ósea en un caso concreto puede dar lugar a numerosas aplicaciones en el campo de las enfermedades óseas, tanto a nivel clínico como experimental. Mediante la combinación del Método de Elementos Finitos (FEM) y los algoritmos de remodelación ósea, es posible obtener modelos numéricos de un hueso específico a partir de datos médicos (por ejemplo, una tomografía computarizada). Todo ello puede dar lugar a una revolución en la medicina personalizada. / [CA] L'accés a una atenció sanitària d'alta qualitat és un marcador important del desenvolupament de les societats humanes. Les aportacions tecnològiques a la medicina han mostrat un potencial rellevant per a descobrir procediments efectius a nivell preventiu, diagnòstic i terapèutic. En particular, els mètodes computacionals permeten el processament eficaç de dades mèdiques i, per tant, poden modelar sistemes biològics complexos. Això ha influït en el desenvolupament de la Medicina Personalitzada (MP) durant les últimes dècades, on l'obtenció de coneixement específic de cada cas permet realitzar intervencions a mesura, tot això a un cost de recursos accessible. La simulació de remodelació òssia és un camp prometedor en el context de la MP. Predir un procés d'adaptació òssia en un cas concret pot donar lloc a nombroses aplicacions en el camp de les malalties òssies, tant a nivell clínic com experimental. Mitjançant la combinació del Mètode d'Elements Finits (*FEM) i els algorismes de remodelació òssia, és possible obtindre models numèrics d'un os específic a partir de dades mèdiques (per exemple, una tomografia computada). Tot això pot donar lloc a una revolució en la medicina personalitzada. / [EN] Access to high-quality healthcare is an important marker of the development of human societies. Technological contributions to medicine have shown relevant potential to discover effective procedures at a preventive, diagnostic and therapeutic level. In particular, computational methods enable efficient processing of medical data and can therefore model complex biological systems. This has influenced the development of Personalized Medicine (PM) over recent decades, where obtaining specific knowledge of each case allows for tailored interventions, all at an affordable resource cost. Simulation of bone remodeling is a promising field in the context of PM. Predicting a bone adaptation process in a specific case can lead to numerous applications in the field of bone diseases, both clinically and experimentally. By combining the Finite Element Method (FEM) and bone remodeling algorithms, it is possible to obtain numerical models of a specific bone from medical data (for example, a CT scan). All of this can lead to a revolution in personalized medicine. / Thanks to the Valencian funding programme FDGENT/2018, for providing economic resources to develop this long-term work. / Gutiérrez Gil, J. (2023). Personalization of Bone Remodelling Simulation Models for Clinical Applications [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/202059