|
| |
|
|
Classification of sleep phases using polysomnographic data
Králík, Martin ; Kozumplík, Jiří (referee) ; Ronzhina, Marina (advisor)
Aim of this thesis is the classification of polysomnographic data. The first part of the thesis is a review of mentioned topic and also the statistical analysis of classification features calculated from real EEG, EOG and EMG for evaluating of the features suitability for sleep stages scoring. The second part is focused on the automatic classification of the data using artificial neural networks. All the results are presented and discussed.
|
|
| |
|
|
Analysis of AVG signals
Musil, Václav ; Sekora, Jiří (referee) ; Rozman, Jiří (advisor)
The presented thesis discusses the basic analysis methods of arteriovelocitograms. The core of this work rests in classification of signals and contribution to possibilities of noninvasive diagnostic methods for evaluation patients with peripheral ischemic occlusive arterial disease. The classification employs multivariate statistical methods and principles of neural networks. The data processing works with an angiographic verified set of arteriovelocitogram dates. The digital subtraction angiography classified them into 3 separable classes in dependence on degree of vascular stenosis. Classification AVG signals are represented in the program by the 6 parameters that are measured on 3 different places on each patient’s leg. Evaluation of disease appeared to be a comprehensive approach at signals acquired from whole patient’s leg. The sensitivity of clustering method compared with angiography is between 82.75 % and 90.90 %, specificity between 80.66 % and 88.88 %. Using neural networks sensitivity is in range of 79.06 % and 96.87 %, specificity is in range of 73.07 % and 91.30 %.
|
|
|
Object Detection and Recognition in Image
Muzikářová, Michaela ; Hradiš, Michal (referee) ; Zemčík, Pavel (advisor)
This bachelor's thesis deals with design and implementation of client-server application for object recognition with the use of existing mobile application. Theoretical part describes the differences between human and computer vision, followed by information about object detection and recognition with selected methods. The next section provides a detailed overview of artificial neural networks, which were used for this work, with their qualities for object recognition. Following part examines selected mobile applications for object recognition, followed by existing frameworks and libraries with focus on artificial neural networks. Among these, Caffe Framework was selected for the work. The next section illustrates the progress of design and implementation and describes the system, along with experiments and dataset used to prove its functionality.
|
|
|
Movement Prediction of Wireless Nodes in Mobile Ad Hoc Networks (MANETS)
Makhlouf, Nermin ; Šimák, Boris (referee) ; Slavíček, Karel (referee) ; Koton, Jaroslav (advisor)
Rychlý vývoj v oblasti mobilní informatiky vyústil v nový, alternativní způsob mobilní komunikace, v němž mobilní uzly tvoří samoorganizující se bezdrátovou síť, jíž se říká mobilní síť ad hoc (Mobile Ad hoc Network, MANET). Specifické vlastnosti sítí MANET stavějí návrh síťového protokolu před řadu problémů na všech vrstvách protokolové sady . Příčinou jsou nepředvídatelné změny topologie a mobilní povaha těchto sítí. Nástrojem, který řeší problémy plynoucí z mobility uzlů, je predikce budoucích změn v topologii sítě. To má zásadní význam pro různé úlohy jako přesměrování. Tato disertační práce se zabývá dvěma metodami predikce mobility pro sítě MANET. První metoda se nazývá „predikce mobility s využitím virtuální mapy“ (mobility prediction using virtual map) a předpokládá, že každý uzel si dokáže vybudovat svou virtuální mapu v závislosti na svém umístění v průběhu času. Vyvinutý predikční algoritmus byl implementován do síťového simulátoru NS-2, aby jej bylo možné vyhodnotit. V této práci zkoumám stávající modely mobility a způsob, jakým v nich lze aplikovat tuto metodu predikce. Simulace sledují zlepšení výkonnosti, co se týče průměrného zpoždění na bázi end-to-end, poměru doručených paketů a propustnosti sítě. Navržený koncept predikce byl implementován pomocí směrovacího protokolu AODV(Ad Hoc On-Demand Distance Vector). Pro druhou metodu jsem vyvinula umělou neuronovou síť pro predikci pohybů v sítích MANET. Model pro predikci mobility vznikl na základě dat shromážděných ze vzorců umístění. K učení či trénování ANN byl využit bayesovský přístup. Ten byl implementován v softwaru pro trénování bayesovských neuronových sítí s názvem Model Manager. Nejlepším způsobem hodnocení závěrečného modelu je provedení predikcí a jejich srovnání s cílovými daty. Predikce vznikají na základě 50 vzorců jako vstupních proměnných. Dosažené výsledky prezentované s diskutované v práci se vyznačují zlepšením zásadních parametrů komunikační sítě, jako jsou propustnost, zpoždění, Poměr doručených paketů, až o 30% v porovnání s klasickým směrovacím protokolem AODV, kde není implementován predikční model.
|
|
|
Classification of ECG by artificial neural networks
Loviška, David ; Vítek, Martin (referee) ; Hrubeš, Jan (advisor)
The aim of project with name Classification ECG by artificial neural networks is simplify and speed up working a doctor. That reaches created program that the is capable simply and almost at once classify EKG signal using artificial neuronal nets. Created program will give to the doctor basic information about used electrocardiogram, as are time period and amplitude signal in single surveyed sections. Subsequently will program warn doctor about abnormalities from normal. Part of program is also graphic window with painted signal and on him in color points and partitions marked by program behind special. In next phase program alone classifies gained data and designating without doctor diagnose that doctor can evaluate and in case agreeable it sign and place for true diagnose patient. This program is also fit for data reading from bigger of the number of hours as far as days. It is concerned primarily Holter ECG monitoring.
|
|
| |
|
|
Word2vec Models with Added Context Information
Šůstek, Martin ; Rozman, Jaroslav (referee) ; Zbořil, František (advisor)
This thesis is concerned with the explanation of the word2vec models. Even though word2vec was introduced recently (2013), many researchers have already tried to extend, understand or at least use the model because it provides surprisingly rich semantic information. This information is encoded in N-dim vector representation and can be recall by performing some operations over the algebra. As an addition, I suggest a model modifications in order to obtain different word representation. To achieve that, I use public picture datasets. This thesis also includes parts dedicated to word2vec extension based on convolution neural network.
|
|
|
Visual detection of small objects using available tools in MATLAB
Sladký, Jiří ; Dobossy, Barnabás (referee) ; Appel, Martin (advisor)
This thesis investigates possibilities of small object detection in pictures using YOLO method, a deep learning algorithm available in MATLAB. In the thesis, a detector was designed and trained to detect cows from top-down view. A tool was created, that performs detection using the proposed model even on high resolution images and counts the present objects. A generator of synthetic images was programmed, which helped with training the model. Various experiments were performed that found the limits of YOLO and validated contribution of the proposed improvements.
|
guest ::
