Wyniki wyszukiwania - BazTech

1

Keystroke dynamics analysis using machine learning methods

Shabliy Nataliya, Lupenko Serhii, Lutsyk Nadiia, Yasniy Oleh, Malyshevska Olha

Applied Computer Science

|

2021

|

Vol. 17, no 4

75--83

EN

The primary objective of the paper was to determine the user based on its keystroke dynamics using the methods of machine learning. Such kind of a problem can be formulated as a classification task. To solve this task, four methods of supervised machine learning were employed, namely, logistic regression, support vector machines, random forest, and neural network. Each of three users typed the same word that had 7 symbols 600 times. The row of the dataset consists of 7 values that are the time period during which the particular key was pressed. The ground truth values are the user id. Before the application of machine learning classification methods, the features were transformed to z-score. The classification metrics were obtained for each applied method. The following parameters were determined: precision, recall, f1-score, support, prediction, and area under the receiver operating characteristic curve (AUC). The obtained AUC score was quite high. The lowest AUC score equal to 0.928 was achieved in the case of linear regression classifier. The highest AUC score was in the case of neural network classifier. The method of support vector machines and random forest showed slightly lower results as compared with neural network method. The same pattern is true for precision, recall and F1-score. Nevertheless, the obtained classification metrics are quite high in every case. Therefore, the methods of machine learning can be efficiently used to classify the user based on keystroke patterns. The most recommended method to solve such kind of a problem is neural network.

2

Implementation of the game model of the Polish Ekstraklasa team using machine learning techniques

Pogodziński Jakub

Annals of Computer Science and Information Systems

|

2021

|

Vol. 26

231

EN

The aim of the thesis is to create a model defining the style of play of a team playing in the Polish Ekstraklasa. The limitation to the highest Polish league class is dictated by the differences in the style of play depending on the league. The model is to be created on the basis of data about the team's game. To build the model, supervised and unsupervised learning techniques will be used and compared to find the relationship between the team's statistics and the determination of its playing style.

3

Enabling Autonomous Medical Image Data Annotation: A human-in-the-loop Reinforcement Learning Approach

da Cruz Leonardo, Sierra-Franco César, Silva-Calpa Greis, Raposo Alberto

Annals of Computer Science and Information Systems

|

2021

|

Vol. 25

271--279

EN

Deep learning techniques have shown significant contributions to several fields, including medical image analysis. For supervised learning tasks, the performance of these techniques depends on a large amount of training data as well as labeled data. However, labeling is an expensive and time-consuming process. With this limitation, we introduce a new approach based on Deep Reinforcement Learning (DRL) to cost-effective annotation in a set of medical data. Our approach consists of a virtual agent to automatically label training data, and a human-in-the-loop to assist in the training of the agent. We implemented the Deep Q-Network algorithm to create the virtual agent and adopted the method mentioned above, which employs human advice to the virtual agent. Our approach was evaluated on a set of medical X-ray data in different use cases, where the agent was required to create new annotations in the form of bounding boxes from unlabeled data. Results show that an agent training with advice positively impacts obtaining new annotations from a data set with scarce labels. This result opens up new possibilities for advancing the study and implementing autonomous approaches with human advice to create a cost-effective annotation in data sets for computer-aided medical image analysis.

4

Basic quantum circuits for classification and approximation tasks

Wiśniewska Joanna, Sawerwain Marek, Obuchowicz Andrzej

International Journal of Applied Mathematics and Computer Science

|

2020

|

Vol. 30, no. 4

733--744

EN

We discuss a quantum circuit construction designed for classification. The circuit is built of regularly placed elementary quantum gates, which implies the simplicity of the presented solution. The realization of the classification task is possible after the procedure of supervised learning which constitutes parameter optimization of Pauli gates. The process of learning can be performed by a physical quantum machine but also by simulation of quantum computation on a classical computer. The parameters of Pauli gates are selected by calculating changes in the gradient for different sets of these parameters. The proposed solution was successfully tested in binary classification and estimation of basic non-linear function values, e.g., the sine, the cosine, and the tangent. In both the cases, the circuit construction uses one or more identical unitary operations, and contains only two qubits and three quantum gates. This simplicity is a great advantage because it enables the practical implementation on quantum machines easily accessible in the nearest future.

5

A semi-automated annotation algorithm based on weakly supervised learning for medical images

Li Hailiang, Zhang Bin, Zhang Yu, Liu Weiwei, Mao Yijun, Huang Jiacheng, Wei Linfeng

Biocybernetics and Biomedical Engineering

|

2020

|

Vol. 40, no. 2

787--802

EN

For medical image recognition, deep learning requires a massive training set, while anno-tation work is a tedious and time-consuming process because of the high technical thresh-old. Furthermore, it is difficult to guarantee annotation accuracy due to the knowledge, skills, and status of the annotator. In this research, we propose a semi-automated annota- tion model based on weakly supervised learning. Moreover, a target-level annotation method is proposed based on weakly supervised learning that is guided by machine learning. The machine learning method is used to screen the regions of interest (RoIs), whose semantic feature vectors are extracted by the deep learning method. Then, the machine learning method is used to cluster them, and the RoIs are finally classified and labeled by a distance comparison. Therefore, this model achieves target-level semi-auto- mated annotation by only using image-level annotations. We applied this method to ultrasound imaging of thyroid papillary carcinoma. The experiments demonstrate the potential of this new methodology to reduce the workload for pathologists and increase the objectivity of diagnoses. We find that 89.8% of papillary thyroid carcinoma regions can be detected automatically, while 82.6% of benign and normal tissue can be excluded without the use of any additional immunohistochemical markers or human intervention.

6

Zastosowanie uczenia nadzorowanego do klasyfikacji defektów PPZ na podstawie sygnałów EA

Wotzka D., Cichoń A., Manowski P.

Poznan University of Technology Academic Journals. Electrical Engineering

|

2018

|

No. 93

335--344

PL

Tematyka artykułu dotyczy rozpoznawania defektów podobciążeniowych przełączników zaczepów (PPZ) z wykorzystaniem uczenia nadzorowanego. PPZ to specjalistyczne urządzenie będące częścią transformatora elektroenergetycznego, które pozwala na skokową zmianę przekładni a tym samym napięcia na zaciskach tego transformatora. Jako metodę diagnostyczną zastosowano metodę emisji akustycznej (EA), której zaletą jest możliwość stosowania podczas normalnej pracy transformatora bez konieczności jego wyłączania. Sygnały EA pozyskane z badań laboratoryjnych, w których symulowano cztery rodzaje defektów - typowych uszkodzeń PPZ, poddano wstępnej analizie z wykorzystaniem filtrów cyfrowych i transformaty Hilberta, a następnie poddano procesowi klasyfikacji. W artykule zawarto przykładowe przebiegi czasowe sygnałów EA oraz wyniki wstępnych badań dotyczących klasyfikacji defektów PPZ z wykorzystaniem siedmiu metod wraz z oceną ich skuteczności.

EN

The subject of the article concerns recognition of defects of on load tap changers (OLTC) with the use of supervised learning. OLTC is a specialized device that is part of a power transformer, which allows for a step change of the gear and thus the voltage at the terminals of this transformer. The acoustic emission (AE) method was applied as diagnostic method. The advantage of this method lies in the possibility of its application during normal operation of the device without having to turn it off. EA signals were obtained from laboratory tests in which four types of defects - typical OLTC damages, were simulated. The gathered signals were pre-analyzed using digital filters and Hilbert transforms, and then subjected to the classification process. The article contains examples of EA signal waveforms and the results of preliminary research on the classification of OLTC defects with the use of seven methods together with an assessment of their effectiveness.

7

Sztuczne sieci neuronowe i kierunki ich zastosowań w przemyśle. Podstawowe rodzaje sztucznych sieci neuronowych

Drobina R., Grecka A., Lewandowski S.

Przegląd Włókienniczy - Włókno, Odzież, Skóra

|

2016

|

nr 7

26--31

EN

The article presents the basic types of artificial neural networks (ANN), designed to solve the regression problems, engineering applications, engineering manufacturing as well as in industrial conditions. The group included these networks are Adaline network, Madaline networks, linear, unidirectional network perpceptron type of multi-layer (MLP), Generalized Regression Neural Networks (GRNN) and a network of radial basis function (RBF).

8

Porównanie skuteczności wybranych klasyﬁkatorów neuronowych w identyﬁkacji grup macerałów węgla kamiennego oraz materii nieorganicznej

Skiba M., Młynarczuk M.

Prace Instytutu Mechaniki Górotworu PAN

|

2016

|

T. 18, nr 4

133--138

PL

Jednym z podstawowych zastosowań sztucznych sieci neuronowych jest rozpoznawanie i klasyfikacja wzorców. W ramach pracy przeprowadzono automatyczną identyfikację grup macerałów oraz materii nieorganicznej za pomocą trzech klasyfikatorów neuronowych: dwuwarstwowej sieci jednokierunkowej (Multi-Layer Perceptron, MLP), sieci o radialnych funkcjach bazowych (Radial Basis Function, RBF) oraz samoorganizującej mapy Kohonena (Self- -Organizing Maps, SOM). Do analiz wykorzystano zbiór 3000 mikroskopowych zdjęć próbek węgla kamiennego. Każde z nich opisano 12 – wymiarowym wektorem cech. Dla każdej z rozpatrywanych sieci dokonano 100 – krotnego powtórzenia losowego wyboru ciągu uczącego, treningu sieci oraz rozpoznania badanych obiektów. Analizy wykazały wysoką skuteczność zastosowanych klasyfikatorów neuronowych w identyfikacji grup macerałów oraz materii nieorganicznej. Najlepsze rezultaty, na poziomie przekraczającym 98% poprawnych rozpoznań, uzyskano dla klasyfikatorów bazujących na uczeniu nadzorowanym (MLP oraz RBF). Nieznacznie niższą skuteczność rozpoznań otrzymano w przypadku sieci SOM – 95,9% klasyfikacji zgodnych z decyzjami obserwatora.

EN

One of the main applications of artificial neural networks is the recognition and classification of different patterns. In the framework of the work an automatic identification of maceral groups and inorganic matter was carried. Three neural classifiers were used: a Multi-Layer Perceptron (MLP), a network of Radial Basis Function (RBF) and Kohonen Self-Organizing Maps (SOM). For the purposes of the analysis a collection of 3,000 images of microscopic samples of coal was used. Each image was described by 12-dimensional feature vector. For each network were carried out: a hundredfold draw of learning set, the network training and classification of objects under study. The analyses have shown high effectiveness of the neural classifi ers used to identify maceral groups and inorganic matter. The best results were obtained for the classifiers based on supervised learning (MLP and RBF). They were at a level exceeding 98% of correct diagnoses. Slightly lower efficiency of diagnosis was obtained in the case of SOM network – 95.9% of classification compatible with the observer decisions.

9

Improving the efficacy of automated fetal state assessment with fuzzy analysis of delivery outcome

Czabanski R., Jezewski M., Horoba K., Jezewski J., Leski J.

Journal of Medical Informatics & Technologies

|

2015

|

Vol. 24

223--230

EN

A number of methods of the qualitative assessment of fetal heart rate (FHR) signals are based on supervised learning. The classification methods based on the supervised learning require a set of training recordings accompanied by the reference interpretation. In the real data collections the class of signals related to fetal distress is usually under-represented. Too small percentage of distress patterns adversely affects the effectiveness of the automated evaluation of the fetal state. The paper presents a method of equalizing the class sizes based on the reference assessment of the fetal state with the fuzzy analysis of the newborn attributes. The supervised learning with increased number of the FHR signals, which are characterized by the highest rate of the fuzzy inference leads to significant increase of the efficacy of the qualitative assessment of the fetal state using the Lagrangian support vector machine.

10

A strategy learning model for autonomous agents based on classification

Śnieżyński B.

International Journal of Applied Mathematics and Computer Science

|

2015

|

Vol. 25, no. 3

471--482

EN

In this paper we propose a strategy learning model for autonomous agents based on classification. In the literature, the most commonly used learning method in agent-based systems is reinforcement learning. In our opinion, classification can be considered a good alternative. This type of supervised learning can be used to generate a classifier that allows the agent to choose an appropriate action for execution. Experimental results show that this model can be successfully applied for strategy generation even if rewards are delayed. We compare the efficiency of the proposed model and reinforcement learning using the farmer–pest domain and configurations of various complexity. In complex environments, supervised learning can improve the performance of agents much faster that reinforcement learning. If an appropriate knowledge representation is used, the learned knowledge may be analyzed by humans, which allows tracking the learning process.

11

Analiza danych niezrównoważonych we wstępnej diagnostyce raka pęcherza moczowego

Piotrowska E., Stanisławski W.

Pomiary Automatyka Kontrola

|

2012

|

R. 58, nr 8

737-740

PL

Artykuł przedstawia wyniki rozważań dotyczących klasyfikacji danych niezrównoważonych w obrazach mikroskopowych preparatów cytologicznych. Do klasyfikacji wykorzystano algorytmy uczenia nadzorowanego jak: naiwny klasyfikator Bayesa, analiza dyskryminacyjna, drzewa decyzyjne oraz zaproponowany przez autorów algorytm klasyfikacji będący połączeniem zbiorów przybliżonych i metody k-najbliższych sąsiadów. Do analizy wykorzystano opracowane przez autorów narzędzie Rough Sets Analysis Toolbox (RSA Toolbox) - przybornik dla środowiska MATLAB. Wykorzystane obrazy mikroskopowe uzyskano w procesie diagnostyki nowotworu pęcherza moczowego badając metodą FISH odpowiednio przygotowane preparaty moczu.

EN

In the paper the results of imbalanced data classification based on microscope images are described. The images were acquired in the process of bladder cancer diagnosis using the FISH method. The conducted research were focused on the effectiveness of the initial cancer diagnosis using specimen radiation in a DAPI channel and supervised learning methods. The analyzed data set contains about 23,000 objects described by 212 morphometric features. Each object was classified to one of two classes: normal cells or cancers cells. Decisions about belonging objects to the corresponding classes were carried out by an expert. There were identified only 640 cancer cells in the analyzed data. Most of learning algorithms assume balance between classes. The class imbalance problem causes difficulties at a learning stage and reduces the predictive ability. Therefore, the classifier evaluation was performed using G-mean and F-value measures. The authors defined additional measure FMaxSen=sen2ospe which is the product of sensitivity and specificity coefficients. Use of the second power factor emphasizes the importance of sensitivity and allows searching the classifier with the maximum specificity at the maximum sensitivity. The analysis presented in the paper was performed with use of Rough Sets Analysis Toolbox (RSA Toolbox) for MATLAB implemented by the authors. The main part of the RSA Toolbox contains a module which supports the rough sets theory processing. Another part (RSAm module) is a wrapper for the proposed rough classification functions and others implemented in Matalab such as NaiveBayes, Discriminant Analysis, Decision Tree. The RSAm gives us possibility to use cross validation for measuring the classification accuracy. The RSAm also contains features reduction algorithms (correlation based feature selection, sequential feature selection, principal component analysis) as well as discretizations algorithms (EWD, CAIM, CACC). An important part of the RSAToolbox is implementation of distributed computations using Matlab Parallel Computing Toolbox and Distributed Computing Server.

12

Multi-label classification using error correcting output codes

Kajdanowicz T., Kazienko P.

International Journal of Applied Mathematics and Computer Science

|

2012

|

Vol. 22, no. 4

829-840

EN

A framework for multi-label classification extended by Error Correcting Output Codes (ECOCs) is introduced and empirically examined in the article. The solution assumes the base multi-label classifiers to be a noisy channel and applies ECOCs in order to recover the classification errors made by individual classifiers. The framework was examined through exhaustive studies over combinations of three distinct classification algorithms and four ECOC methods employed in the multi-label classification problem. The experimental results revealed that (i) the Bode-Chaudhuri-Hocquenghem (BCH) code matched with any multi-label classifier results in better classification quality; (ii) the accuracy of the binary relevance classification method strongly depends on the coding scheme; (iii) the label power-set and the RAkEL classifier consume the same time for computation irrespective of the coding utilized; (iv) in general, they are not suitable for ECOCs because they are not capable to benefit from ECOC correcting abilities; (v) the all-pairs code combined with binary relevance is not suitable for datasets with larger label sets.

13

Human identification based on a kinematical data of a gait

Świtoński A., Mucha R., Danowski D., Mucha M., Cieślar G., Wojciechowski K., Sieroń A.

Przegląd Elektrotechniczny

|

2011

|

R. 87, nr 12b

169--172

EN

The paper is devoted to the gait identification challenges. It evaluates human abilities to recognize gait on the basis of skeleton animations. Further, it proposes the method of gait identification based on the kinematical data .The feature extraction approach and supervised learning are applied. To explore the most individual joints movements, aggregated feature rankings are calculated. To examine the proposed method, the database containing 353 gaits of 25 different actors is collected in the motion capture laboratory. We have obtained 99.7% of classification accuracy.

PL

W artykule zaprezentowano eksperyment oceniający zdolności człowieka do rozpoznawania chodu oraz zaproponowano metodę identyfikacji chodu na podstawie danych kinematycznych. Bazuje ona na podejściu z ekstrakcją cech i uczeniem nadzorowanym. W celu oceny ruchu poszczególnych stawów pod kątem osobniczych cech różnicujących wyznaczono zagregowane rankingi atrybutów. Do weryfikacji zaproponowanej metody, zgromadzono bazę 353 przejść wykonywanych przez 25 różnych aktorów. Uzyskano ponad 99% skuteczność klasyfikacji.

14

Diagnosis of the motion pathologies based on a reduced kinematical data of a gait

Świtoński A., Mucha R., Danowski D., Mucha M., Polański A., Cieślar G., Wojciechowski K., Sieroń A.

Przegląd Elektrotechniczny

|

2011

|

R. 87, nr 12b

173--176

EN

The paper presents method of motion analysis supporting diagnosis of gait abnormalities on the basis of reduced kinematical data of a gait. The proposed method consist of the following steps: kinematical data reduction by Principal Component Analysis, determination of the Fourier component for the 3D PCA trajectories and supervised learning. To examine proposed approach, we have collected database of gaits containing data of coxarthrosis patients. We have got 100% of classification accuracy for the considered disease.

PL

W artykule zaprezentowano metodę analizy danych ruchu dla celów wspierania diagnostyki nieprawidłowości chodu. W kolejnych krokach przeprowadzana jest redukcja wymiarowości kinematycznych danych chodu z wykorzystaniem metody analizy składowych głównych, wyznaczane są składowe Fouriera dla otrzymanych trójwymiarowych trajektorii PCA oraz przeprowadzane jest uczenie nadzorowane. W celu weryfikacji zaproponowanej metody zgromadzono bazę danych przejść ze schorzeniami stawu biodrowego, dla której to udało się uzyskać 100% skuteczność klasyfikacji.

15

Analysis of the resume learning process for spiking neural networks

Ponulak F.

International Journal of Applied Mathematics and Computer Science

|

2008

|

Vol. 18, no 2

117-127

EN

In this paper we perform an analysis of the learning process with the ReSuMe method and spiking neural networks (Ponulak, 2005; Ponulak, 2006b). We investigate how the particular parameters of the learning algorithm affect the process of learning. We consider the issue of speeding up the adaptation process, while maintaining the stability of the optimal solution. This is an important issue in many real-life tasks where the neural networks are applied and where the fast learning convergence is highly desirable.

16

A comparative review of the selection methods for discovering differentially expressed genes in microarray experiments for classification

Stąpor K., Bruckner A., Błaszczyk P.

Studia Informatica

|

2006

|

Vol. 27, nr 4

37-52

EN

In this paper the feature selection methods applied to discovering differentially expressed genes in microarray experiments are compared. This compare-son includes both filter and optimal subset selection methods. The simulated and biological datasets are used as the microarray gene expression data, and the ability of selected genes for classification is also considered.

PL

W artykule porównano metody selekcji cech zastosowane do wykrywania genów różnicujących w eksperymentach mikromacierzowych. Porównanie zawiera zarówno metody statystyczne, jak i metody poszukiwania optymalnego podzbioru cech. Jako dane mikromacierzowe wykorzystano symulowane zbiory danych oraz dane biologiczne. Przedstawiono ponadto przydatność wyselekcjonowanych genów do klasyfikacji.

17

Comparison of supervised learning methods for spike time coding in spiking neural networks

Kasiński A., Ponulak F.

International Journal of Applied Mathematics and Computer Science

|

2006

|

Vol. 16, no 1

101-113

EN

In this review we focus our attention on supervised learning methods for spike time coding in Spiking Neural Networks (SNNs). This study is motivated by recent experimental results regarding information coding in biological neural systems, which suggest that precise timing of individual spikes may be essential for efficient computation in the brain. We are concerned with the fundamental question: What paradigms of neural temporal coding can be implemented with the recent learning methods? In order to answer this question, we discuss various approaches to the learning task considered. We shortly describe the particular learning algorithms and report the results of experiments. Finally, we discuss the properties, assumptions and limitations of each method. We complete this review with a comprehensive list of pointers to the literature.

18

Nadzorowana detekcja tras komunikacyjnych z wykorzystaniem metod uczenia maszynowego

Krawiec K., Wyczałek I.

Archiwum Fotogrametrii, Kartografii i Teledetekcji

|

2006

|

Vol. 16

361--371

PL

W pracy opisujemy nową metodę wykrywania drogowych tras komunikacyjnych na zobrazowaniu lotniczym lub satelitarnym. Proponowana metoda ma charakter strukturalny i bazuje na koncepcji profilu, rozumianego jako lokalny jednowymiarowy przekrój (rzut) obrazu. Tak rozumiane profile podlegają analizie poprzez ekstrakcję z nich cech zorientowanych na dyskryminowanie punktów reprezentujących drogi od punktów reprezentujących inne obiekty widoczne w obrazie. Cechy analizowane w proponowanej metodzie dobrane zostały do charakterystyki szlaków komunikacyjnych (głównie podłużny kształt); należą do nich m.in. wzajemne podobieństwo blisko zlokalizowanych profili o tej samej orientacji (ciągłość) oraz symetria. Dla polepszenia precyzji, profile obliczane są z wykorzystaniem próbkowania podpunktowego (sub-pixel sampling). W dalszych etapach przetwarzania metoda wykorzystuje algorytmy uczenia maszynowego (machine learning), w szczególności nadzorowane uczenie się z przykładów. Algorytm uczący się z przykładów dysponuje uczącą próbką pikseli, dla których przynależność do klas decyzyjnych (droga, nie-droga) jest znana. Informacja ta może być wprowadzona przez decydenta (eksperta) poprzez zaznaczenie wybranego obszaru obrazu reprezentującego szlak komunikacyjny, lub pochodzić z odpowiedniego modułu systemu informacji przestrzennej. Algorytm uczenia maszynowego pozyskuje wiedzę ze zbioru uczącego w procesie uczenia indukcyjnego. Wiedza ta jest następnie stosowana do klasyfikowania pozostałych punktów obrazu, dla których informacja ucząca nie jest znana. Ponadto, ponieważ wiedza ta jest wyrażona w dogodnej postaci drzewa decyzyjnego, może być poddana analizie przez eksperta (i potencjalnie skorygowana). Poza prezentacją metody praca zawiera opis jej implementacji komputerowej oraz eksperymentu obliczeniowego przeprowadzonego na rzeczywistym zdjęciu lotniczym terenu zabudowanego. Otrzymane wyniki dowodzą skuteczności proponowanego algorytmu i wskazują na użyteczność podejścia wykorzystującego uczenie maszynowe do analizy zdjęć lotniczych i obrazów satelitarnych.

EN

This paper presents a novel method of road detection in aerial and satellite imaging. This structural method is based on the concept of profile, meant as a local one-dimensional cross-section (cast) of raster image. We acquire such profiles from the image at different orientation angles and extract from them features well discriminating road pixels from non-road pixels. In particular, we use feature definitions tailored to road characteristics (mostly elongation); these include, among others, mutual similarity of close and equally orientated profiles (road continuity) and symmetry. To improve the precision of analysis, the method computes profiles using sub-pixel sampling. The further part of processing relies on machine learning, in particular, on supervised learning from examples. The algorithm is given a training sample of pixels, for which the decision class assignment (road, non-road) is known. This information may be manually entered by a decision maker (expert) by marking image regions representing road fragments, or alternatively, it may be retrieved from an appropriate module of a geographical information system. Given that information, the algorithm acquires the knowledge from training examples, performing so-called “inductive” learning. That knowledge may be then used to classify the remaining image pixels, for which the decision class assignment is not known. Moreover, the knowledge may be inspected (and potentially corrected) by the decision maker, as it is expressed in a readable form of a decision tree. The paper presents the algorithm in detail, describes its computer implementation, and demonstrates its application to an aerial image of urban area. The obtained results demonstrate the good performance of the method and indicate the usefulness of machine learning approach in analysis of aerial and satellite imagery.