Browsing by Subject "machine learning"

Now showing 1 - 20 of 85

Access status: Restricted ,
A machine learning aided schedule planning system
(Data obrony: 2017-07-05) Koleżyńska, Katarzyna
Wydział Elektrotechniki, Automatyki, Informatyki i Inżynierii Biomedycznej
Access status: Restricted ,
A method of interpretation of vector representation in text analysis
(Data obrony: 2019-09-27) Bigaj, Michał
Wydział Informatyki, Elektroniki i Telekomunikacji
Access status: Open Access ,
A Python library for the Jupyteo IDE Earth observation processing tool enabling interoperability with the QGIS System for use in data science
(Wydawnictwa AGH, 2022) Bednarczyk, Michał
This paper describes JupyQgis - a new Python library for Jupyteo IDE enabling interoperability with the QGIS system. Jupyteo is an online integrated development environment for earth observation data processing and is available on a cloud platform. It is targeted at remote sensing experts, scientists and users who can develop the Jupyter notebook by reusing embedded open-source tools, WPS interfaces and existing notebooks. In recent years, there has been an increasing popularity of data science methods that have become the focus of many organizations. Many scientific disciplines are facing a significant transformation due to data-driven solutions. This is especially true of geodesy, environmental sciences, and Earth sciences, where large data sets, such as Earth observation satellite data (EO data) and GIS data are used. The previous experience in using Jupyteo, both among the users of this platform and its creators, indicates the need to supplement its functionality with GIS analytical tools. This study analyzed the most efficient way to combine the functionality of the QGIS system with the functionality of the Jupyteo platform in one tool. It was found that the most suitable solution is to create a custom library providing an API for collaboration between both environments. The resulting library makes the work much easier and simplifies the source code of the created Python scripts. The functionality of the developed solution was illustrated with a test use case.
Access status: Open Access ,
A universal convolutional neural network for the pixel-level detection and monitoring of weld beads
(Wydawnictwa AGH, 2024) Wang, Zhuo; Kayitmazbatir, Metin; Banu, Mihaela
In weld-based manufacturing processes such as welding and metal deposition additive manufacturing (AM), the weld bead is a direct indicator of manufacturing quality. For example, the geometry of the weld bead was optimized to a net shape which outperformed conventional geometries. Automatic monitoring of weld bead is thus of prime importance for welding process control and quality assurance. This paper develops a general-purpose convolutional neural network (CNN) for pixel-level detection and monitoring of beads, regardless of welding materials, machine, manufacturing conditions, etc. To achieve the generality, we collected a great variety of welding images containing 2677 single-line beads from 231 research articles, followed by pixel-wise hand-annotation. Consequently, the trained CNN can recognize different beads from various backgrounds at a pixel level. Case studies show that compared to the image-level classification in prior research, its pixel-level labeling permits real-time, complete characterization of weld beads (e.g., detailed morphology, discontinuity, spatter, and uniformity) for more informed process control. This research represents a significant step towards developing a truly human-like monitoring system with low-level scene understanding ability and general applicability.
Access status: Open Access ,
Adapting text categorization for manifest based android malware detection
(Wydawnictwa AGH, 2019) Çoban, Önder; Özel, Selma Ayşe
Malware is a shorthand of malicious software that are created with the intent of damaging hardware systems, stealing data, and causing a mess to make money, protest something, or even make war between governments. Malware is often spread by downloading some applications for your hardware from some download platforms. It is highly probable to face with a malware while you try to load some applications for your smart phones nowadays. Therefore it is very important that some tools are needed to detect malware before loading them to the hardware systems. There are mainly three different approaches to detect malware: i) static, ii) dynamic, and iii) hybrid. Static approach analyzes the suspicious program without executing it. Dynamic approach, on the other hand, executes the program in a controlled environment and obtains information from operating system during runtime. Hybrid approach, as its name implies, is the combination of these two approaches. Although static approach may seem to have some disadvantages, it is highly preferred because of its lower cost. In this paper, our aim is to develop a static malware detection system by using text categorization techniques. To reach our goal, we apply text mining techniques like feature extraction by using bag-of-words, n-grams, etc. from manifest content of suspicious programs, then apply text classification methods to detect malware. Our experimental results revealed that our approach is capable of detecting malicious applications with an accuracy between 94.0% and 99.3%.
Access status: Open Access ,
An advanced ensemble modeling approach for predicting carbonate reservoir porosity from seismic attributes
(Wydawnictwa AGH, 2023) Topór, Tomasz; Sowiżdżał, Krzysztof
This study uses a machine learning (ML) ensemble modeling approach to predict porosity from multiple seismic attributes in one of the most promising Main Dolomite hydrocarbon reservoirs in NW Poland. The presented workflow tests five different model types of varying complexity: K-nearest neighbors (KNN), random forests (RF), extreme gradient boosting (XGB), support vector machine (SVM), single layer neural network with multilayer perceptron (MLP). The selected models are additionally run with different configurations originating from the pre-processing stage, including Yeo–Johnson transformation (YJ) and principal component analysis (PCA). The race ANOVA method across resample data is used to tune the best hyperparameters for each model. The model candidates and the role of different pre-processors are evaluated based on standard ML metrics – coefficient of determination (R2), root mean squared error (RMSE), and mean absolute error (MAE). The model stacking is performed on five model candidates: two KNN, two XGB, and one SVM PCA with a marginal role. The results of the ensemble model showed superior accuracy over single learners, with all metrics (R2 0.890, RMSE 0.0252, MAE 0.168). It also turned out to be almost three times better than the neural net (NN) results obtained from commercial software on the same testing set (R2 0.318, RMSE 0.0628, MAE 0.0487). The spatial distribution of porosity from the ensemble model indicated areas of good reservoir properties that overlap with hydrocarbon production fields. This observation completes the evaluation of the ensemble technique results from model metrics. Overall, the proposed solution is a promising tool for better porosity prediction and understanding of heterogeneous carbonate reservoirs from multiple seismic attributes.
Access status: Open Access ,
An empirical analysis of changes in the Błędów Desert using machine learning methods
(Wydawnictwa AGH, 2025) Czernik, Anna; Borowiec, Natalia; Marmol, Urszula
The aim of the study was to determine changes in the land cover of the Błędów Desert, which is a habitat for rare flora and fauna species protected under the Natura 2000 program. Invasive plants, which pose a threat to protected species, are present in this area. Additionally, human activities can have negative impacts on the desert ecosystem. Therefore, the land manager is obligated to carry out actions aimed at maintaining the appropriate size and character of the desert. The analysis was conducted using satellite imagery from the Sentinel-2 mission, which provides images with high temporal and spatial resolution. The study covered the years 2015–2022 and took into account seasonal variability due to the presence of green vegetation. Change detection methods based on data integration, including photointerpretation and machine learning classification, were used for land cover analysis. Five representative land cover classes were defined, enabling a quantitative presentation of changes in the Błędów Desert and a qualitative assessment of the classification performed. The results of the study indicate variability in land cover depending on the season, with an increasing number of protected plant species, including grasslands. Simultaneously, a slight increase in the desert area was noted, manifesting as an increase in sand in forested areas. The results obtained demonstrate the effective implementation of the Natura 2000 program objectives.
Access status: Open Access ,
An evaluation of some machine learning algorithms as tools for predicting soil characteristics based on their spectral response in the Vis‑NIR range
(Wydawnictwa AGH, 2021) Gruszczyński, Stanisław
Using the Land Use and Coverage Frame Survey (LUCAS) database of European soil surface layer properties, statistical and machine learning predictive models for several key soil characteristics (clay content, pH in CaCl2, concentration of organic carbon, calcium carbonates and nitrogen and exchange cations capacity) were compared on the basis of processing their spectral responses in the visible (Vis) and near‑infrared (NIR) parts. Standard methods of relationship modeling were used: stepwise regression, partial least squares regression and linear regression with input data obtained from principal components analysis. Using the inputs extracted by statistical algorithms various machine learning algorithms were used in the modeling. The usefulness of the models was analyzed by comparison with the values of the determination coefficients, the root mean square error and the distribution of residual values. The mean square error of estimation in the cross‑validation procedure for the stack model using the multilayer perceptron and the distributed random forest were as follows: for clay content - ca. 4.5%, for pH - ca. 0.35, for SOC - ca. 7.5 g/kg (0.75% by weight), for CaCO3 content - ca. 19 g/kg, for N content - ca. 0.50 g/kg, and for CEC - ca. 3.5 cmol(+)/kg.
Access status: Open Access ,
Analityka kulturowa, czyli jak narzędzia »data science« zmieniły humanistykę
(Wydawnictwa AGH, 2022) Radomski, Andrzej
The article presents research paradigms that have radically changed the contemporary humanities. The most important of these is cultural analytics. It is based on Data Science methods. The author presents the assumptions of data science, and then the characteristics of digital humanities and cultural analytics. The second part of the article presents examples of research and projects conducted as part of cultural analysis. These are projects implemented at the DH Lab at Yale University, Software Studies Initiative, and Media Lab Katowice. Research conducted in these institutions transformed the humanities. Its characteristic features are the study of large data collections, research automation, the use of machine learning and knowledge visualization. The new humanities, the author claims, has become an exact science.
Access status: Restricted ,
Analiza i klasyfikacja sygnałów EEG związanych z ruchem prawą ręką, lewą ręką, językiem oraz stopą
(Data obrony: 2018-01-22) Król, Alicja
Wydział Elektrotechniki, Automatyki, Informatyki i Inżynierii Biomedycznej
Access status: Restricted ,
Analiza metod maszynowego uczenia do rozpoznawania obrazu
(Data obrony: 2018-01-30) Kostuch, Aleksander
Wydział Elektrotechniki, Automatyki, Informatyki i Inżynierii Biomedycznej
Access status: Restricted ,
Analiza mozliwosci wykorzystania machine learning w przewidywaniu halasu drogowego
(Data obrony: 2021-01-26) Hanczewski, Mikołaj
Wydział Inżynierii Mechanicznej i Robotyki
Access status: Restricted ,
Aplikacja mobilna do predykcji emocji na podstawie charakterystyk głosu
(Data obrony: 2019-01-29) Jakubowski, Marcin
Wydział Elektrotechniki, Automatyki, Informatyki i Inżynierii Biomedycznej
Access status: Restricted ,
Aplikacja webowa do zarządzania procesem uczenia maszynowego
(Data obrony: 2020-01-15) Kamiński, Marcin
Wydział Elektrotechniki, Automatyki, Informatyki i Inżynierii Biomedycznej
Access status: Open Access ,
Application of Basic Machine-Learning Classifiers for Automatic Anomaly Detection in Shewhart Control Charts
(Wydawnictwa AGH, 2024) Woźniak, Aleksander; Krawiec, Klaudia; Książek, Roger
In today’s dynamic technological environment, innovation plays a crucial role – especially for manufacturing enterprises that constantly strive to improve the quality of their products. This article examines the quality-management issue in a company producing car rims. It was identified that real-time quality control can sometimes be unreliable due to controller fatigue, leading to erroneous data interpretation or delayed responses to deviations in the production process. The study aimed to investigate the possibility of eliminating or significantly reducing these errors by employing a tool that is based on artificial intelligence. The article covers the preparation of training data, the training of classifiers, and the evaluation of their effectiveness in analyzing control charts in real time. The adopted hypothesis assumes that machine-learning classifiers can be effective methods of support for quality controllers. The research began with collecting measurement data from the machine and dividing it into training and test sets. The obtained results were evaluated using standard quality measures for machine-learning models. The results showed that the use of artificial intelligence can bring significant benefits in improving quality supervision in the production process of car rims.
Access status: Open Access ,
Application of linguistic cues in the analysis of language of hate groups
(Wydawnictwa AGH, 2015) Balcerzak, Bartłomiej; Jaworski, Wojciech
Hate speech and fringe ideologies are social phenomena that thrive on-line. Members of the political and religious fringe are able to propagate their ideas via the Internet with less effort than in traditional media. In this article, we attempt to use linguistic cues such as the occurrence of certain parts of speech in order to distinguish the language of fringe groups from strictly informative sources. The aim of this research is to provide a preliminary model for identifying deceptive materials online. Examples of these would include aggressive marketing and hate speech. For the sake of this paper, we aim to focus on the political aspect. Our research has shown that information about sentence length and the occurrence of adjectives and adverbs can provide information for the identification of differences between the language of fringe political groups and mainstream media.
Access status: Restricted ,
Application of machine learning algorithms in analysis and prediction of Steam service users behavior
(Data obrony: 2018-01-24) Bizoń, Paweł
Wydział Elektrotechniki, Automatyki, Informatyki i Inżynierii Biomedycznej
Access status: Open Access ,
Artificial intelligence approaches to determine graphite nodularity in ductile iron
(AGH University of Science and Technology Press, 2021) Brait, Maximilian; Koppensteiner, Eduard; Schindelbacher, Gerhard; Li, Jiehua; Schumacher, Peter
The complex metallurgical interrelationships in the production of ductile cast iron can lead to enormous differences in graphite formation and local microstructure by small variations during production. Artificial intelligence algorithms were used to describe graphite formation, which is influenced by a variety of metallurgical parameters. Moreover, complex physical relationships in the formation of graphite morphology are also controlled by boundary conditions of processing, the effect of which can hardly be assessed in everyday foundry operations. The influence of relevant input parameters can be predetermined using artificial intelligence based on conditions and patterns that occur simultaneously. By predicting the local graphite formation, measures to stabilise production were defined and thereby the accuracy of structure simulations improved. In course of this work, the most important dominating variables, from initial charging to final casting, were compiled and analysed with the help of statistical regression methods to predict the nodularity of graphite spheres. We compared the accuracy of the prediction by using Linear Regression, Gaussian Process Regression, Regression Trees, Boosted Trees, Support Vector Machines, Shallow Neural Networks and Deep Neural Networks. As input parameters we used 45 characteristics of the production process consisting of the basic information including the composition of the charge, the overheating time, the type of melting vessel, the type of the inoculant, the fading, and the solidification time. Additionally, the data of several thermal analysis, oxygen activity measurements and the final chemical analysis were included. Initial programme designs using machine learning algorithms based on neural networks achieved encouraging results. To improve the degree of accuracy, this algorithm was subsequently adapted and refined for the nodularity of graphite.
Access status: Open Access ,
Assessment of approaches for the extraction of building footprints from pléiades images
(Wydawnictwa AGH, 2021) Taha, Lamyaa Gamal El-deen; Ibrahim, Rania Elsayed
The Marina area represents an official new gateway of entry to Egypt and the development of infrastructure is proceeding rapidly in this region. The objective of this research is to obtain building data by means of automated extraction from Pléiades satellite images. This is due to the need for efficient mapping and updating of geodatabases for urban planning and touristic development. It compares the performance of random forest algorithm to other classifiers like maximum likelihood, support vector machines, and backpropagation neural networks over the well-organized buildings which appeared in the satellite images. Images were subsequently classified into two classes: buildings and non-buildings. In addition, basic morphological operations such as opening and closing were used to enhance the smoothness and connectedness of the classified imagery. The overall accuracy for random forest, maximum likelihood, support vector machines, and backpropagation were 97%, 95%, 93% and 92% respectively. It was found that random forest was the best option, followed by maximum likelihood, while the least effective was the backpropagation neural network. The completeness and correctness of the detected buildings were evaluated. Experiments confirmed that the four classification methods can effectively and accurately detect 100% of buildings from very high-resolution images. It is encouraged to use machine learning algorithms for object detection and extraction from very high-resolution images.
Access status: Restricted ,
Assessment of the importance of news articles based on the machine learning methods
(Data obrony: 2019-09-27) Głąb, Katarzyna Anna
Wydział Informatyki, Elektroniki i Telekomunikacji