|
|
Reordering Text Fragments Using a Language Model
Holubec, Michael ; Kocour, Martin (referee) ; Beneš, Karel (advisor)
The aim of this work is to construct and experimentally verify the effectiveness of the language model in identifying the reading order. For this purpose language model with LSTM architecture was constructed. This work designs and implements three methods which are used to identify reading order. These methods are Language analysis, Spatial analysis and Combined analysis. Language analysis and combined analysis used constructed language model. The success of the language model, and all three methods, was measured on three datasets containing newspaper articles. Language analysis reaches 57,6 % and spatial analysis reaches 91,6 %. Combined analysis achieved the best results 92,9 %. The work shows that the language model can be used to identify reading order but use of additional data (e.g. spatial data
|
|
|
Artificial Intelligence for a Board Game
Tureček, Dominik ; Baskar, Murali Karthick (referee) ; Beneš, Karel (advisor)
This work proposes and implements AI agents for the game Dice Wars. Dice Wars is turn-based, zero-sum game with non-deterministic move results. Several AI agents were created using rule-based approach, expectiminimax algorithm, and logistic regression. To evaluate the performance of proposed agents, an implementation of the game was created. Results of the experiments have shown that it's preferable to play aggressively in two-player games and make more optimal moves in games played with more players. The agent using expectiminimax is able to win more than 60 % of games in 8-player games against random players and wins 21.4 % of games played against a mix of seven other agents created in this work. In two-player setups, the agent using logistic regression with numbers of players' scores and number of dice as features has the best performance and wins 59.4 % of games in average.
|
|
|
Neural Language Model Acceleration
Labaš, Dominik ; Černocký, Jan (referee) ; Beneš, Karel (advisor)
This work adresses the topic of neural language model acceleration. The aim of this work is to optimize model of a feed-forward neural network. In accelerating of the neural network we used a change of activation function, pre-calculation of matrices for calculationg the hidden layer, implementation of the model's history cache and unnormalized model. The best-performing model was accelerated by 75.3\%.
|
|
|
Acoustic Scene Classification from Speech
Grepl, Filip ; Beneš, Karel (referee) ; Matějka, Pavel (advisor)
This thesis deals with creating a system whose task is to recognize what type of location the recording was created at by analyzing the audio signal. The classifier is based on a multi-layer, fully connected neural network. The topology of the neural network is based on the baseline system provided for the DCASE competition. A dataset from this competition is also used for training and evaluating the neural network. The experiments are performed in particular with the representation of the properties of the audio records and with the format of the input data of the neural network. For this purpose, Mel-filter bank, block Mel-filter bank and MFCC flags are used. The experiments performed in this thesis brought a classification accuracy increased by 6.5 % compared to the baseline system of DCASE. Overall system success rate reached 67.5 %.
|
|
| |
|
|
Detection of Malicious Websites using Machine Learning
Šulák, Ladislav ; Černocký, Jan (referee) ; Beneš, Karel (advisor)
Táto práca sa zaoberá problematikou škodlivého kódu na webe so zameraním na analýzu a detekciu škodlivého JavaScriptu umiestneného na strane klienta s využitím strojového učenia. Navrhnutý prístup využíva známe i nové pozorovania s ohľadom na rozdiely medzi škodlivými a legitímnymi vzorkami. Tento prístup má potenciál detekovať nové exploity i zero-day útoky. Systém pre takúto detekciu bol implementovaný a využíva modely strojového učenia. Výkon modelov bol evaluovaný pomocou F1-skóre na základe niekoľkých experimentov. Použitie rozhodovacích stromov sa podľa experimentov ukázalo ako najefektívnejšia možnosť. Najefektívnejším modelom sa ukázal byť Adaboost klasifikátor s dosiahnutým F1-skóre až 99.16 %. Tento model pracoval s 200 inštanciami randomizovaného rozhodovacieho stromu založeného na algoritme Extra-Trees. Viacvrstvový perceptrón bol druhým najlepším modelom s dosiahnutým F1-skóre 97.94 %.
|
|
| |
|
|
Linear Logistic Regression Demo
Bak, Adam ; Kesiraju, Santosh (referee) ; Beneš, Karel (advisor)
This bachelor's thesis deals with the machine learning model logistic regression.The aim is to closely inspect and analyze the workings of this model for classification, in order to be able to provide a learning tool in the form of demonstrative application. All of the mathematical formulae, logistic sigmoid, cross entropy error function and gradient are derived and explained in a concise manner. This thesis also provides some insight into the form of the cross entropy error function in the case of linear logistic regression.
|
|
|
Machine Translation Using Artificial Neural Networks
Holcner, Jonáš ; Beneš, Karel (referee) ; Szőke, Igor (advisor)
The goal of this thesis is to describe and build a system for neural machine translation. System is built with recurrent neural networks - encoder-decoder architecture in particular. The result is a nmt library used to conduct experiments with different model parameters. Results of the experiments are compared with system built with the statistical tool Moses.
|
|
|
Gaussian Processes Based Hyper-Optimization of Neural Networks
Coufal, Martin ; Landini, Federico Nicolás (referee) ; Beneš, Karel (advisor)
Cílem této diplomové práce je vytvoření nástroje pro optimalizaci hyper-parametrů umělých neuronových sítí. Tento nástroj musí být schopen optimalizovat více hyper-parametrů, které mohou být navíc i korelovány. Tento problém jsem vyřešil implmentací optimalizátoru, který využívá Gaussovské procesy k predikci vlivu jednotlivých hyperparametrů na výslednou přesnost neuronové sítě. Z provedených experimentů na několika benchmark funkcích jsem zjistil, že implementovaný nástroj je schopen dosáhnout lepších výsledků než optimalizátory založené na náhodném prohledávání a snížit tak v průměru počet potřebných kroků optimalizace. Optimalizace založená na náhodném prohledávání dosáhla lepších výsledků pouze v prvních krocích optimalizace, než si optimalizátor založený na Gaussovských procesech vytvoří dostatečně přesný model problému. Nicméně téměř všechny experimenty provedené na datasetu MNIST prokázaly lepší výsledky optimalizátoru založeného na náhodném prohledávání. Tyto rozdíly v provedených experimentech jsou pravděpodobně dány složitostí zvolených benchmark funkcí nebo zvolenými parametry implementovaného optimalizátoru.
|
guest ::
RSS feed.