|
|
Reservoir Computing for Industrial Applications
Brhel, Jakub ; Bražina, Jakub (referee) ; Kovář, Jiří (advisor)
In this bachelor thesis, the field of reservoir computing and its application in the industrial sector is investigated. The main objective of the thesis is to verify the effectiveness and accuracy of the echo state network method in time series prediction. To achieve this objective, a dataset from a manufacturing machine is used and the echo state network algorithm is implemented. The results are analysed and are also compared with the results of previous studies in the field of reservoir computing.
|
|
|
EEG Classification Model for Emotion Detection Using Python
Vengerová, Veronika ; Zaheer, Muhammad Asad (referee) ; Jawed, Soyiba (advisor)
Táto práca sa zaoberá rozoznávaním emócií z elektroencefalogramu (EEG). Dva modely na binárnu klasifikáciu emócií, kde jeden model klasifikuje neutrálnu emóciu alebo strach a druhý šťastie a smútok. Počas práce boli vyskúšané mnohé rôzne architektúry, pričom najlepšie výsledky boli dosiahnuté modelom pozostávajúcim z dvoch vetiev KNN-LSTM spojenými pred výstupnou vrstvou. Výsledná presnosť bola 87.309% na klasifikáciu šťastia a smútku a 84.865% na klasifikáciu neutrálnej emócie a strachu.
|
|
|
Time-Series Analysis and Prediction by Means of Neural Networks
Kňažovič, Martin ; Jaroš, Jiří (referee) ; Bidlo, Michal (advisor)
This thesis deals with stock price prediction based on the creation of prediction models for selected stocks (BRK-A, GOOG, and MSFT), which can help investors in the creation of their financial decisions or by replacing other stock prediction models in existing prediction systems. Models created in this thesis are presented in two types - univariate model and multivariate model, which are in their final version presented in two architectures, one-layer architecture and two-layer architecture. Discussed models are created by means of neural networks, specifically recurrent neural networks with its extension - Long short-term memory. The output of the presented models is a forecast of the next-day stock price, which can be used for evaluating the right time to buy or sell a given stock. The quality of individual prediction models is evaluated via the mean squared error of the validation or testing dataset or alternatively based on stock price trend prediction.
|
|
|
Improving Consistency in Text Recognition Datasets
Tvarožný, Matúš ; Hradiš, Michal (referee) ; Kišš, Martin (advisor)
This work is concerned with increasing the consistency of datasets for text recognition. This paper describes the problems that cause the inconsistency and then presents solutions to eliminate it. The effect of the properties of the polygons defining the text line boundaries and hence how the modified version of the dataset, which is composed of ideal text line variants, affected the accuracy of the model is investigated. Further, the work focuses on detecting and then removing or modifying text lines whose ground truth transcription does not match the actual text they contain. Experimentation showed that removing the visual inconsistency on the training set did not have a significant effect on the trained model, but modifying the test set improved the OCR accuracy of the model by 1.1\% CER. By modifying the dataset so that it did not contain mutually inconsistent pairs of recognized text and the corresponding ground truth, the model improved by a maximum of only 0.2\% CER after re-training. The main finding of this work is, above all, the proven beneficial effect of removing inconsistencies on test suites, thanks to which it is possible to determine a more realistic error rate of the OCR model.
|
|
|
Web application for Cybersecurity Job Ads Analysis
Turek, Adam ; Sikora, Marek (referee) ; Ricci, Sara (advisor)
Cílem bakalářské práce je vytvoření interaktivní celosvětové mapy zobrazující databázi pracovních inzerátů ve webové aplikaci a provedení filtrování podle různých parametrů, kde je následně provedena analýza strojového učení. Také mapa zobrazuje počet inzerátů na pracovní pozice podle příslušných států. Webová aplikace je vytvořena pomoci JavaScriptové knihovny ReactJS spojené s LeafletJS, které zajišťují hlavní funkcionalitu. Část se strojovým učením a změna skriptů je realizována pomocí programovacího jazyku Python. Práce popisuje teoretickou část a implementaci jednotlivých funkcí mapy a dále se zabývá popisem a úspěsnou úpravou skriptů pro účely provedení strojového učení.
|
|
|
Neural Network Based Named Entity Recognition
Straková, Jana ; Hajič, Jan (advisor) ; Černocký, Jan (referee) ; Konopík, Miloslav (referee)
Title: Neural Network Based Named Entity Recognition Author: Jana Straková Institute: Institute of Formal and Applied Linguistics Supervisor of the doctoral thesis: prof. RNDr. Jan Hajič, Dr., Institute of Formal and Applied Linguistics Abstract: Czech named entity recognition (the task of automatic identification and classification of proper names in text, such as names of people, locations and organizations) has become a well-established field since the publication of the Czech Named Entity Corpus (CNEC). This doctoral thesis presents the author's research of named entity recognition, mainly in the Czech language. It presents work and research carried out during CNEC publication and its evaluation. It fur- ther envelops the author's research results, which improved Czech state-of-the-art results in named entity recognition in recent years, with special focus on artificial neural network based solutions. Starting with a simple feed-forward neural net- work with softmax output layer, with a standard set of classification features for the task, the thesis presents methodology and results, which were later used in open-source software solution for named entity recognition, NameTag. The thesis finalizes with a recurrent neural network based recognizer with word embeddings and character-level word embeddings,...
|
|
| |
|
|
Automatic Harmony Generation
Bobčík, Martin ; Drahošová, Michaela (referee) ; Vašíček, Zdeněk (advisor)
Goal of this master thesis is to study harmonization based on knowledge of given melody and to design a system which will meaningfully automate this activity. In the work there is covered basics of music theory needed for this topic and previous other approaches to this problematic. There is also covered machine learning, neural networks and recurrent neural networks. In the end, there is outlined design of the system, how to make it work and how to use it. Four experiments were executed with the system. Harmonization of the short melodies were unpleasant. Harmonization of longer melodies were overall more successful though. A possible cause might be relatively small used neural network of the system.
|
|
|
Automatic Harmony Generation
Bobčík, Martin ; Drahošová, Michaela (referee) ; Vašíček, Zdeněk (advisor)
Goal of this master thesis is to study harmonization based on knowledge of given melody and to design a system which will meaningfully automate this activity. In the work there is covered basics of music theory needed for this topic and previous other approaches to this problematic. There is also covered machine learning, neural networks and recurrent neural networks. In the end, there is outlined design of the system, how to make it work and how to use it. Three experiments were executed with the system. Harmonization of the melodies were unpleasant though. A possible cause might be relatively small used neural network of the system.
|
|
|
Personal Voice Activity Detection
Sedláček, Šimon ; Landini, Federico Nicolás (referee) ; Švec, Ján (advisor)
Cílem této práce je implementovat, otestovat a vyhodnotit řečníkem podmíněnou metodu pro detekci hlasu ( Voice Activity Detection , VAD) nazvanou Personal VAD. Pro detekci využívá tato metoda LSTM neuronových sítí a jejím účelem je vytvoření systému schopného spolehlivě detekovat řečové signály cílového řečníka při zachování vlastností typického VAD systému co se velikosti modelu, odezvy a nízkých nároků na zdroje týče. Systém je trénován pro klasifikaci řečových rámců do tří tříd: neřeč, řeč necílového a řeč cílového řečníka. Za tímto účelem využívá metoda speaker embedding vektory pro reprezentaci cílového řečníka jako součást vstupních příznaků. Některé z náročnějších variant systému využívají skórování rámců systémem pro verifikaci řečníka, což vede ke zvýšení spolehlivosti klasifikace. Vedle základní metody skórování představené v originálním článku byly navrženy dvě modifikace, jež základní metodu překonaly a zlepšily spolehlivost výsledného systému i v akusticky náročných prostředích.
|
guest ::
