|
|
Detection of Anomalies in Pedestrian Walking
Pokorný, Ondřej ; Orság, Filip (referee) ; Goldmann, Tomáš (advisor)
The goal of this work was to create a system that would be able to detect anomalies in pedestrian walking. As the core of my application, I have used OpenPose, which is an application for detecting human skeletons. Then I used a bidirectional LSTM neural network to detect anomalies in video sequences. This architecture was chosen during the experiment because it outperformed other solutions. I trained my model to detect three types of anomalies. The output of my application is a video with marked sequences of anomalies. The whole system is implemented in Python.
|
|
|
Detection of Unusual Events in Temporal Data
Černík, Tomáš ; Bartík, Vladimír (referee) ; Zendulka, Jaroslav (advisor)
Bachelor thesis deals with detection of unusual events (anomalies) in available temporal data. Theoretical part describes existing techniques and algorithms used to detect outliers. There are also introduced meteorological data that are after that used for experimental verification of implemented detection algorithms. Second part, practical one, describes design and implementation of application and algorithms. Algorithms are also tested in search for point, contextual and collective anomalies.
|
|
|
Statistical anomaly detection methods of data communication
Woidig, Eduard ; Mangová, Marie (referee) ; Slavíček, Karel (advisor)
This thesis serves as a theoretical basis for a practical solution to the issue of the use of statistical methods for detecting anomalies in data traffic. The basic focus of anomaly detection data traffic is on the data attacks. Therefore, the main focus is the analysis of data attacks. Within the solving are data attacks sorted by protocols that attackers exploit for their own activities. Each section describes the protocol itself, its usage and behavior. For each protocol is gradually solved description of the attacks, including the methodology leading to the attack and penalties on an already compromised system or station. For the most serious attacks are outlined procedures for the detection and the potential defenses against them. These findings are summarized in the theoretical analysis, which should serve as a starting point for the practical part, which will be the analysis of real data traffic. The practical part is divided into several sections. The first of these describes the procedures for obtaining and preparing the samples to allow them to carry out further analysis. Further described herein are created scripts that are used for obtaining needed data from the recorded samples. These data are were analyzed in detail, using statistical methods such as time series and descriptive statistics. Subsequently acquired properties and monitored behavior is verified using artificial and real attacks, which is the original clean operation modified. Using a new analysis of the modified traffics compared with the original samples and an evaluation of whether it has been some kind of anomaly detected. The results and tracking are collectively summarized and evaluated in a separate chapter with a description of possible further attacks, which were not directly part of the test analysis.
|
|
|
Machine Learning from Intrusion Detection Systems
Dostál, Michal ; Očenášek, Pavel (referee) ; Hranický, Radek (advisor)
The current state of intrusion detection tools is insufficient because they often operate based on static rules and fail to leverage the potential of artificial intelligence. The aim of this work is to enhance the open-source tool Snort with the capability to detect malicious network traffic using machine learning. To achieve a robust classifier, useful features of network traffic were choosed, extracted from the output data of the Snort application. Subsequently, these traffic features were enriched and labeled with corresponding events. Experiments demonstrate excellent results not only in classification accuracy on test data but also in processing speed. The proposed approach and the conducted experiments indicate that this new method could exhibit promising performance even when dealing with real-world data.
|
|
|
Application Firewall Anomaly Detection
Pospěch, Jan ; Homoliak, Ivan (referee) ; Očenášek, Pavel (advisor)
The goal of the presented bachelor thesis is to describe the process of anomaly detection in application firewalls. The thesis focuses on the principles and basics of anomaly detection, the reader is introduced to the techniques and methods of machine learning. The process of analyzing the requests and responses received from the web application protection system is described, and the system design is developed. The practical part describes the implementation of the system and testing on real datasets. Decision tree and Random forest algorithms show the best results with f1-score 0.9987. Among the unsupervised learning methods, the best results are shown by Autoencoder with an f1-score value of 0.8315.
|
|
|
Anomaly Recognition in Advanced Detection Systems
Poposki, Vasil ; Homoliak, Ivan (referee) ; Očenášek, Pavel (advisor)
Cílem této práce je implementovat systém detekce anomálií využívající techniky umělé inteligence, který dokáže detekovat anomálie učením chování systému. Navrhovaný přístup je účinný při identifikaci nových nebo neznámých anomálií, které tradiční metody založené na pravidlech mohou postrádat v datech síťového provozu. Implementace takového systému však zahrnuje i řešení problémů, jako je zpracování dat a extrakce charakteristických rysů. Tato práce pojednává o různých metodách analýzy dat a přístupech k odhalení průniků v systémech Extended Detection and Response a výzvách, kterým čelíme v dnešních rozšiřujících se bezpečnostních technologiích.
|
|
|
UI for the Cooperation of a Surveillance System with a Human
Klem, Václav ; Špaňhel, Jakub (referee) ; Bažout, David (advisor)
This bachelor's thesis discusses the topic of the design and development of an effective user interface for the cooperation of a user and a surveillance system. It offers a study of currently available tools for the annotation of videos and detection/tracking of people in these videos to create an annotation application intended for new way of detecting anomalies. This innovative way is based on detecting anomalies on the level of frames. Anomaly in this sense means any illegal behavior of people in a video. This thesis aims to create a proposition of UI design for the annotation application, confirm the validity of the concept and then implement the full-fledged version of this application based on the testing results. During the testing of the prototype, the annotation success rate of 89% and the rating of overall UI clarity of 83% respectively were reached. Following the results of the testing, the final version of the application was implemented reflecting the user feedback.
|
|
|
Malformations/anomalies in the development of cestodes
Aliaskerova, Madina ; Schreiber, Manfred (advisor) ; Chanová, Marta (referee)
Anomalies in tapeworms can appear spontaneously under natural conditions, at both morphological and developmental level. In adult tapeworms, malformations at morphological level are manifested on the scolex and strobila. Their detailed description can be found in Taenia saginata, Taenia pisiformis, Taenia solium, Dibothriocephalus nihonkaiense, Dibothriocephalus latus, Hymenolepis nana and Hymenolepis microstoma. In the larval stages of Taenia crassiceps, the malformations occur predominantly on the skolex. Morphological malformations may be manifested by multiplication of suckers or change in the appearance and number of hooks, change in the structure of the genitalia, occurrence of lateral segments, fenestration of the strobila or occurrence of multiple planes of symmetry. However, these malformations can also be caused by targeted radiation exposure, temperature stress or anthelmintics. Developmental anomalies are also manifested by spontaneous appearance of tapeworms in different parts of the body in different host species. There is a link between anomalous infections and the immune status of the host. Possible causes of anomalies include damage to neoblasts, lack of a proper immune response by the host, anthelmintics, host diet, or environmental influences.
|
|
|
Capital Market Anomalies
ALEŠ, Petr
This Master thesis deals with the anomalies in capital markets. Through statistical testing of data from five companies on the US stock exchange NASDAQ seeks to prove or disprove their presence on this market.
|
|
|
Mispricing in leveraged value small-capitalization stocks
Picálek, Jan ; Hronec, Martin (advisor) ; Novák, Jiří (referee)
We study returns in the universe of leveraged value small-capitalization stocks, a universe with historically significant exposure to common risk factors. We sep- arate future winners and losers within this universe of risky stocks by adopting machine-learning-based mispricing strategy. The strategy considers 34 stock- level characteristics to predict 1-month-ahead returns and construct a long- short portfolio accordingly. The portfolio yields abnormal risk-adjusted re- turns of 0.42% per month out-of-sample, uncovering statistically significant mispricing. The machine-learning algorithm is trained on leveraged value small- capitalization stocks, so it captures universe-specific nonlinearities and variable interactions. The nonlinear effects and predictive power of individual variables are extracted and presented as well. We found no evidence of a relationship between the magnitude of the mispricing and credit cycles, or market volatility. JEL Classification G11, G12, G14, Keywords Anomalies, Predictability of returns, Asset pricing tests, Leveraged equities, Value stocks Title Mispricing in leveraged value small-capitalization stocks
|
guest ::
