|
|
Surrogate-assisted Evolutionary Computation
Konečný, Štěpán ; Dobrovský, Ladislav (referee) ; Kůdela, Jakub (advisor)
This thesis deals with the optimization of flight profile design using surrogate models with differential evolution. The aim was to become familiar with the problem and possible solution procedures. The main objective was to create a surrogate model, implement it on the problem and then optimize it. In the introduction of the paper, the used surrogate models and evolutionary algorithms are described in detail. Subsequently, the control method, the optimization of expensive problems, the data acquisition issues for surrogate models and examples of practical applications of surrogate models are described. The paper concludes with a description of the optimization procedure and the results obtained by optimizing the flight profile design using the selected surrogate models, followed by an evaluation.
|
|
|
Benchmarking of evolutionary algorithms
Kostiha, Petr ; Matoušek, Radomil (referee) ; Kůdela, Jakub (advisor)
This thesis focuses on the benchmarking of surrogate-assisted evolutionary algorithms, an area that has not yet been explored as deeply as traditional evolutionary methods. The aim is to identify and analyze the best available algorithms of this type and highlight their potential applications. The paper begins with the definition of an optimization problem, followed by a theoretical overview of benchmarking, its significance, and tools. Subsequently, selected evolutionary methods are detailed and then implemented in MATLAB software for the experimental part of the work. Out of eight tested methods, the SAMSO method proved to be the most effective, along with the LSADE, ESA, and TS-DDEO methods. The results show that these methods offer significant potential for solving complex optimization problems and open up new possibilities for further research in this field.
|
|
|
Torque-speed characteristic optimization of an induction machine using machine learning
Bártková, Tereza ; Klíma, Petr (referee) ; Bílek, Vladimír (advisor)
Purpose of this work is optimization of given electrical machine based on combination of selected methods. The first chapter introduces machine learning in general. Its fundamental approaches and typical problems are analyzed together with some of the actual algorithms. The topic of optimization is introduced in the following chapter - its purpose and explanation of its concepts. The limitations we face while optimizing designs are demonstrated on two basic methods. Several popular algorithms are described, that could possibly be used in the context of electrical machines' optimization. One of the goals is reached in the third chapter - the choice of machine learning method in combination with optimization algorithm. The fourth chapter deals with matters of creating the induction machine geometry and subsequently its electromagnetic model for the following analysis. Details of geometric dimensions and electromagnetic parameters are introduced as well as calculation of end winding leakage inductance, one phase winding resistance and resistance and inductance between bars. There is also a brief description of the tools used for creating the model. The next chapter presents results of the electromagnetic analysis of the model. The process of creation on surrogate models of a given machine is described in the sixth chapter. It comprises the sensitivity analysis, creation of the initial training data and active learning. Next chapter deals with the optimization of the given machine utilizing surrogate models and selected optimization algorithms. The last chapter compares original machines' characteristics to those of optimized geometries.
|
|
|
Surrogate-assisted Evolutionary Computation
Konečný, Štěpán ; Dobrovský, Ladislav (referee) ; Kůdela, Jakub (advisor)
This thesis deals with the optimization of flight profile design using surrogate models with differential evolution. The aim was to become familiar with the problem and possible solution procedures. The main objective was to create a surrogate model, implement it on the problem and then optimize it. In the introduction of the paper, the used surrogate models and evolutionary algorithms are described in detail. Subsequently, the control method, the optimization of expensive problems, the data acquisition issues for surrogate models and examples of practical applications of surrogate models are described. The paper concludes with a description of the optimization procedure and the results obtained by optimizing the flight profile design using the selected surrogate models, followed by an evaluation.
|
|
|
Evolutionary Algorithms for Multiobjective Optimization
Pilát, Martin ; Neruda, Roman (advisor) ; Schoenauer, Marc (referee) ; Pošík, Petr (referee)
Multi-objective evolutionary algorithms have gained a lot of atten- tion in the recent years. They have proven to be among the best multi-objective optimizers and have been used in many industrial ap- plications. However, their usability is hindered by the large number of evaluations of the objective functions they require. These can be expensive when solving practical tasks. In order to reduce the num- ber of objective function evaluations, surrogate models can be used. These are a simple and fast approximations of the real objectives. In this work we present the results of research made between the years 2009 and 2013. We present a multi-objective evolutionary algo- rithm with aggregate surrogate model, its newer version, which also uses a surrogate model for the pre-selection of individuals. In the next part we discuss the problem of selection of a particular type of model. We show which characteristics of the various models are im- portant and desirable and provide a framework which combines sur- rogate modeling with meta-learning. Finally, in the last part, we ap- ply multi-objective optimization to the problem of hyper-parameters tuning. We show that additional objectives can make finding of good parameters for classifiers faster. 1
|
|
|
Evolutionary algorithms and active learning
Repický, Jakub ; Holeňa, Martin (advisor) ; Fink, Jiří (referee)
Názov práce: Evoluční algoritmy a aktivní učení Autor: Jakub Repický Katedra: Katedra teoretické informatiky a matematické logiky Vedúci diplomovej práce: doc. RNDr. Ing. Martin Holeňa, CSc., Ústav informa- tiky, Akademie věd České republiky Abstrakt: Vyhodnotenie ciel'ovej funkcie v úlohách spojitej optimalizácie často do- minuje výpočtovej náročnosti algoritmu. Platí to najmä v prípade black-box fun- kcií, t. j. funkcií, ktorých analytický popis nie je známy a ktoré sú vyhodnocované empiricky. Témou urýchl'ovania black-box optimalizácie s pomocou náhradných modelov ciel'ovej funkcie sa zaoberá vel'a autorov a autoriek. Ciel'om tejto dip- lomovej práce je vyhodnotit' niekol'ko metód, ktoré prepájajú náhradné modely založené na Gaussovských procesoch (GP) s Evolučnou stratégiou adaptácie ko- variančnej matice (CMA-ES). Gaussovské procesy umožňujú aktívne učenie, pri ktorom sú body pre vyhodnotenie vyberané s ciel'om zlepšit' presnost' modelu. Tradičné náhradné modely založené na GP zah'rňajú Metamodelom asistovanú evolučnú stratégiu (MA-ES) a Optimalizačnú procedúru pomocou Gaussovských procesov (GPOP). Pre účely tejto práce boli oba prístupy znovu implementované a po prvý krát vyhodnotené na frameworku Black-Box...
|
|
|
Evolutionary Algorithms for Multiobjective Optimization
Pilát, Martin ; Neruda, Roman (advisor) ; Schoenauer, Marc (referee) ; Pošík, Petr (referee)
Multi-objective evolutionary algorithms have gained a lot of atten- tion in the recent years. They have proven to be among the best multi-objective optimizers and have been used in many industrial ap- plications. However, their usability is hindered by the large number of evaluations of the objective functions they require. These can be expensive when solving practical tasks. In order to reduce the num- ber of objective function evaluations, surrogate models can be used. These are a simple and fast approximations of the real objectives. In this work we present the results of research made between the years 2009 and 2013. We present a multi-objective evolutionary algo- rithm with aggregate surrogate model, its newer version, which also uses a surrogate model for the pre-selection of individuals. In the next part we discuss the problem of selection of a particular type of model. We show which characteristics of the various models are im- portant and desirable and provide a framework which combines sur- rogate modeling with meta-learning. Finally, in the last part, we ap- ply multi-objective optimization to the problem of hyper-parameters tuning. We show that additional objectives can make finding of good parameters for classifiers faster. 1
|
guest ::
