Iskanje po katalogu digitalne knjižnice

Opis: This master thesis is focusing on the utilization of nature-inspired algorithms for hyperparameter optimization, how they work and how to use them. We present some existing methods for hyperparameter optimization as well as propose a novel method that is based on six different nature-inspired algorithms: Firefly algorithm, Grey Wolf Optimizer, Particle Swarm Optimization, Genetic algorithm, Differential Evolution, and Hybrid Bat algorithm. We also show the optimization results (set of hyperparameters) for each algorithm and we present the plots of the accuracy for each combination and handpicked one. In discussion of the results, we provide the answers on our research questions as well as propose ideas for future work.
Ključne besede: artificial intelligence, artificial neural networks, machine learning, nature-inspired algorithms, evolutionary algorithms
Objavljeno v DKUM: 09.12.2019; Ogledov: 2232; Prenosov: 122
Celotno besedilo (969,13 KB)

Opis: In some patients with ventricular fibrillation (VF) there may be a better chance of successful defibrillation after a period of chest compression and ventilation before the defibrillation attempt. It is therefore important to know whether a defibrillation attempt will be successful. The predictive powerof a model developed by "genetic" programming (GP) to predict defibrillation success was studied. Methods and Results: 203 defibrillations were administered in 47 patients with out-of-hospital cardiac arrest due to a cardiac cause. Maximal amplitude, a total energy of power spectral density, and the Hurst exponent of the VF electrocardiogram (ECG) signal were included in the model developed by GP. Positive and negative likelihood ratios of the model for testing data were 35.5 and 0.00, respectively. Using a model developed by GP on the complete database, 120 of the 124 unsuccessful defibrillations would have been avoided, whereas all of the 79 successful defibrillations would have been administered. Conclusion: The VF ECG contains information predictive of defibrillation success. The model developed by GP, including data from the time-domain, frequency-domain and nonlinear dynamics, could reduce the incidence of unsuccessful defibrillations.
Ključne besede: optimisation methods, evolutionary optimisation methods, genetic algorithms, genetic programming, defibrillation, cardiac arrest prediction
Objavljeno v DKUM: 01.06.2012; Ogledov: 2046; Prenosov: 104
Povezava na celotno besedilo

Opis: Atrial fibrillation (AF) is the most common rhythm disorder. Because of the high recurrence rate of AF after cardioversion and because of potential side effects of electrical cardioversion, it is clinically important to predict persistence of sinus rhythm after electrical cardioversion before it is attempted. The aim of our study was the development of a mathematical model by"genetic" programming (GP), a non-deterministic modelling technique, which would predict maintenance of sinus rhythm after electrical cardioversion of persistent AF. PATIENTS AND METHODS: Ninety-seven patients with persistent AF lasting more than 48 h, undergoing the first attempt at transthoracic cardioversion were included in this prospective study. Persistence of AF before the cardioversion attempt, amiodarone treatment, left atrial dimension,mean, standard deviation and approximate entropy of ECG R-R intervals were collected. The data of 53 patients were randomly selected from the database and used for GP modelling; the other 44 data sets were used for model testing. RESULTS: In 23 patients sinus rhythm persisted at 3 months. In the other 21 patients sinus rhythm was not achieved or its duration was less than 3 months. The model developed by GP failed to predict maintenance ofsinus rhythm at 3 months in one patient and in six patients falsely predicted maintenance of sinus rhythm. Positive and negative likelihood ratiosof the model for testing data were 4.32 and 0.05, respectively. Using this model 15 of 21 (71.4%) cardioversions not resulting in sinus rhythm at 3 months would have been avoided, whereas 22 of 23 (95.6%) cardioversions resulting in sinus rhythm at 3 months would have been administered. CONCLUSION: This model developed by GP, including clinical data, ECG data from the time-domain and nonlinear dynamics can predict maintenance of sinus rhythm. Further research is needed to explore its utility in the present or anexpanded form.
Ključne besede: optimisation methods, evolutionary optimisation methods, genetic algorithms, genetic programming, defibrillation, cardiac arrest prediction, atrial fibrillation, electrical cardioversion, prediction
Objavljeno v DKUM: 01.06.2012; Ogledov: 2352; Prenosov: 84
Povezava na celotno besedilo

Ključne besede: hybridization, evolutionary algorithms, heuristic algorithms
Objavljeno v DKUM: 01.06.2012; Ogledov: 2339; Prenosov: 83
Povezava na celotno besedilo