Search
Search results for: EXPLAINABLE CLASSIFIERS, COUNTERFACTUAL APPROACH, BIAS DETECTION
-
TOWARDS EXPLAINABLE CLASSIFIERS USING THE COUNTERFACTUAL APPROACH - GLOBAL EXPLANATIONS FOR DISCOVERING BIAS IN DATA
PublicationThe paper proposes summarized attribution-based post-hoc explanations for the detection and identification of bias in data. A global explanation is proposed, and a step-by-step framework on how to detect and test bias is introduced. Since removing unwanted bias is often a complicated and tremendous task, it is automatically inserted, instead. Then, the bias is evaluated with the proposed counterfactual approach. The obtained results...
-
Application of a stochastic compartmental model to approach the spread of environmental events with climatic bias
PublicationWildfires have significant impacts on both environment and economy, so understanding their behaviour is crucial for the planning and allocation of firefighting resources. Since forest fire management is of great concern, there has been an increasing demand for computationally efficient and accurate prediction models. In order to address this challenge, this work proposes applying a parameterised stochastic model to study the propagation...
-
Direct spectrum detection based on Bayesian approach
PublicationThe paper investigates the Bayesian framework's performance for a direct detection of spectrum parameters from the compressive measurements. The reconstruction signal stage is eliminated in by the Bayesian Compressive Sensing algorithm, which causes that the computational complexity and processing time are extremely reduced. The computational efficiency of the presented procedure is significantly...
-
An A-Team Approach to Learning Classifiers from Distributed Data Sources
Publication -
An A-Team approach to learning classifiers from distributed data sources
Publication -
Playback detection using machine learning with spectrogram features approach
PublicationThis paper presents 2D image processing approach to playback detection in automatic speaker verification (ASV) systems using spectrograms as speech signal representation. Three feature extraction and classification methods: histograms of oriented gradients (HOG) with support vector machines (SVM), HAAR wavelets with AdaBoost classifier and deep convolutional neural networks (CNN) were compared on different data partitions in respect...
-
An Approach to the Detection of Bank Robbery Acts Employing Thermal Image Analysis
PublicationA novel approach to the detection of selected security-related events in bank monitoring systems is presented. Thermal camera images are used for the detection of people in difficult lighting conditions. Next, the algorithm analyses movement of objects detected in thermal or standard monitoring cameras using a method evolved from the motion history images algorithm. At the same time, thermal images are analyzed in order to detect...
-
MultiRegional PCA for leakage detection and localisation in DWDS - approach
PublicationMonitoring is one of the most important parts in advanced control of complex dynamic systems. Information about systems behavior, including failures indicating, enables for efficient control. The chapter describes an approach to detection and localisation of pipe leakage in Drinking Water Distribution Systems (DWDS) representing complex and distributed dynamic system of large scale. Proposed MultiRegional Principal Component Analysis...
-
Compressive Sensing Approach to Harmonics Detection in the Ship Electrical Network
PublicationThe contribution of this paper is to show the opportunities for using the compressive sensing (CS) technique for detecting harmonics in a frequency sparse signal. The signal in a ship’s electrical network, polluted by harmonic distortions, can be modeled as a superposition of a small number of sinusoids and the discrete Fourier transform (DFT) basis forms its sparse domain. According to the theory of CS, a signal may be reconstructed...
-
Real-time PCR approach in dermatophyte detection and Trichophyton rubrum identification
PublicationDermatophytes are keratinophilic molds that infect human hair, nails and skin. Diagnosis of dermatophytosis is based on morphological, serological and biochemical features. However, identification is difficult and laborious due to similarities between microorganisms. Thus, there is considerable interest to develop mycological diagnostic procedures based on molecular biology methods. In this study, fast, two-step DNA extraction...
-
Gender approach to multi-objective optimization of detection systems by pre-selection of criteria
PublicationA novel idea of performing evolutionary computations for solving highly-dimensional multi-objective optimization (MOO) problems is proposed. The information about individual genders is applied. This information is drawn out of the fitness of individuals and applied during the parental crossover in the evolutionary multi-objective optimization (EMO) processes. The paper introduces the principles of the genetic-gender approach (GGA)...
-
Gender approach to multi-objective optimization of detection systems by pre-selection of criteria
PublicationA novel idea of performing evolutionary computations for solving highly-dimensional multi-objective optimization (MOO) problems is proposed. The information about individual genders is applied. This information is drawn out of the fitness of individuals and applied during the parental crossover in the evolutionary multi-objective optimization (EMO) processes. The paper introduces the principles of the genetic-gender approach (GGA)...
-
Object-oriented approach to oil spill detection using ENVISAT ASAR images
Publication -
Cointegration approach for temperature effect compensation in Lamb-wave-based damage detection
Publication -
Gender approach to multi-objective optimization of detection systems by pre-selection of criteria
PublicationA novel idea of performing evolutionary computations for solving highly-dimensional multi-objective optimization (MOO) problems is proposed. The information about individual genders is applied. This information is drawn out of the fitness of individuals and applied during the parental crossover in the evolutionary multi-objective optimization (EMO) processes. The paper introduces the principles of the genetic-gender approach (GGA)...
-
Multifrequency Nanoscale Impedance Microscopy (m-NIM): A novel approach towards detection of selective and subtle modifications on the surface of polycrystalline boron-doped diamond electrodes
PublicationIn this paper, we describe the modification of Nanoscale Impedance Microscopy (NIM), namely, a combination of contact-mode atomic force microscopy with local impedance measurements. The postulated approach is based on the application of multifrequency voltage perturbation instead of standard frequency-by-frequency analysis, which among others offers more time-efficient and accurate determination of the resultant impedance spectra...
-
Cointegration and wavelet-analysis-based approach for Lamb-wave-based structural damage detection
Publication -
Stationarity-Based Approach for the Selection of Lag Length in Cointegration Analysis Used for Structural Damage Detection
Publication -
Orientation-aware ship detection via a rotation feature decoupling supported deep learning approach
PublicationShip imaging position plays an important role in visual navigation, and thus significant focuses have been paid to accurately extract ship imaging positions in maritime videos. Previous studies are mainly conducted in the horizontal ship detection manner from maritime image sequences. This can lead to unsatisfied ship detection performance due to that some background pixels maybe wrongly identified as ship contours. To address...
-
A novel approach exploiting properties of convolutional neural networks for vessel movement anomaly detection and classification
PublicationThe article concerns the automation of vessel movement anomaly detection for maritime and coastal traffic safety services. Deep Learning techniques, specifically Convolutional Neural Networks (CNNs), were used to solve this problem. Three variants of the datasets, containing samples of vessel traffic routes in relation to the prohibited area in the form of a grayscale image, were generated. 1458 convolutional neural networks with...
-
Fast-response optoelectronic detection of explosives’ residues from the nitroaromatic compounds detonation: field studies approach
PublicationWe are presenting an application of optoelectronic nitrogen dioxide (NO2) analyzer based on cavity enhanced absorption spectroscopy in the detection of traces of explosives after detonation. It has been shown that the analyzer using blue-violet laser is able to detect explosive residues after the detonation of various amounts of nitroaromatic compounds (75g-1kg) with higher efficiency than the HPLC soil sample testing equipment,...
-
Wioleta Kucharska dr hab. inż.
PeopleWioleta Kucharska holds a position as an Associate Professor at the Faculty of Management and Economics of the Gdansk TECH, Gdansk University of Technology, Fahrenheit Universities Union, Poland. She authored 74 peer-reviewed studies published with Wiley, Springer, Taylor & Francis, Emerald, Elsevier, IGI Global, Routledge, and by members of international conferences committees. She reviews international journals (83 revisions...
-
Automatic labeling of traffic sound recordings using autoencoder-derived features
PublicationAn approach to detection of events occurring in road traffic using autoencoders is presented. Extensions of existing algorithms of acoustic road events detection employing Mel Frequency Cepstral Coefficients combined with classifiers based on k nearest neighbors, Support Vector Machines, and random forests are used. In our research, the acoustic signal gathered from the microphone placed near the road is split into frames and converted...
-
What should we know when choosing feather, blood, egg or preen oil as biological samples for contaminants detection? A non-lethal approach to bird sampling for PCBs, OCPs, PBDEs and PFASs
PublicationBirds are considered as good bio-monitors and they can provide highly valuable data about the level of contamination in their habitat. During the design of biomonitoring studies one of the first issues after choosing species is the choice of biological material. Non-lethally collected samples have recently been gaining greater attention as they offer several ethical and practical advantages. However, not all sample matrices are...
-
Deep Learning-Based, Multiclass Approach to Cancer Classification on Liquid Biopsy Data
PublicationThe field of cancer diagnostics has been revolutionized by liquid biopsies, which offer a bridge between laboratory research and clinical settings. These tests are less invasive than traditional biopsies and more convenient than routine imaging methods. Liquid biopsies allow studying of tumor-derived markers in bodily fluids, enabling the development of more precise cancer diagnostic tests for screening, disease monitoring, and...
-
Zdzisław Kowalczuk prof. dr hab. inż.
PeopleZdzislaw Kowalczuk received his M.Sc. degree in 1978 and Ph.D. degree in 1986, both in Automatic Control from Technical University of Gdańsk (TUG), Gdańsk, Poland. In 1993 he received his D.Sc. degree (Dr Habilitus) in Automatic Control from Silesian Technical University, Gliwice, Poland, and the title of Professor from the President of Poland in 2003. Since 1978 he has been with Faculty of Electronics, Telecommunications and Informatics...
-
Empirical Analysis of Forest Penalizing Attribute and Its Enhanced Variations for Android Malware Detection
PublicationAs a result of the rapid advancement of mobile and internet technology, a plethora of new mobile security risks has recently emerged. Many techniques have been developed to address the risks associated with Android malware. The most extensively used method for identifying Android malware is signature-based detection. The drawback of this method, however, is that it is unable to detect unknown malware. As a consequence of this problem,...
-
Adding Interpretability to Neural Knowledge DNA
PublicationThis paper proposes a novel approach that adds the interpretability to Neural Knowledge DNA (NK-DNA) via generating a decision tree. The NK-DNA is a promising knowledge representation approach for acquiring, storing, sharing, and reusing knowledge among machines and computing systems. We introduce the decision tree-based generative method for knowledge extraction and representation to make the NK-DNA more explainable. We examine...
-
PCA based Fault Tolerant MPC
PublicationThis chapter presents a Fault Tolerant - Model Predictive Control (FT-MPC) schemes for sensor faults accommodation. A Fault Detection and Isolation (FDI) Unit, which is an integral part of FT-MPC system, is based on the Principal Component Analysis (PCA) method. Introduced approach enables efficient bias and drift faults accommodation in single, as well as simultaneous faults case. Simple simulation exercise is presented.Rozdział...
-
Improving automatic surveillance by sound analysis
PublicationAn automatic surveillance system, based on event detection in the video image can be improved by implementing algorithms for audio analysis. Dangerous or illegal actions are often connected with distinctive sound events like screams or sudden bursts of energy. A method for detection and classification of alarming sound events is presented. Detection is based on the observation of sudden changes in sound level in distinctive sub-bands...
-
Detection of anomalies in bee colony using transitioning state and contrastive autoencoders
PublicationHoneybees plays vital role for the environmental sustainability and overall agricultural economy. Assisting bee colonies within their proper functioning brings the attention of researchers around the world. Electronics systems and machine learning algorithms are being developed for classifying specific undesirable bee behaviors in order to alert about upcoming substantial losses. However, classifiers could be impaired when used...
-
Buzz-based honeybee colony fingerprint
PublicationNon-intrusive remote monitoring has its applications in a variety of areas. For industrial surveillance case, devices are capable of detecting anomalies that may threaten machine operation. Similarly, agricultural monitoring devices are used to supervise livestock or provide higher yields. Modern IoT devices are often coupled with Machine Learning models, which provide valuable insights into device operation. However, the data...
-
Diagnostic Models and Estimators for LDI in Transmission Pipelines
PublicationThis article considers and compares four analytical models of the pipeline flow process for leak detection and location tasks. The synthesis of these models is briefly outlined. Next, the methodology for generating data and diagnosing pipes is described, as well as experimental settings, assumptions and implemented scenarios. Finally, the quality of model-based diagnostic estimators has been evaluated for their bias, standard deviations...
-
Multiscaled Hybrid Features Generation for AdaBoost Object Detection
PublicationThis work presents the multiscaled version of modified census features in graphical objects detection with AdaBoost cascade training algorithm. Several experiments with face detector training process demonstrate better performance of such features over ordinal census and Haar-like approaches. The possibilities to join multiscaled census and Haar features in single hybrid cascade of strong classifiers are also elaborated and tested....
-
Electrical responses of Graphene-Silicon Schottky diodes toward nitrogen dioxide and tetrahydrofuran under irradiation
Open Research DataGraphene-Silicon Schottky junctions were utilized as gas sensors toward inorganic (nitrogen dioxide) and organic (tetrahydrofuran) gas qualitative and quantitative detection. The electrical responses of the sensors were collected in the form of current-voltage characteristics and measurements of current in time domain for a selected voltage bias. The...
-
Study of Multi-Class Classification Algorithms’ Performance on Highly Imbalanced Network Intrusion Datasets
PublicationThis paper is devoted to the problem of class imbalance in machine learning, focusing on the intrusion detection of rare classes in computer networks. The problem of class imbalance occurs when one class heavily outnumbers examples from the other classes. In this paper, we are particularly interested in classifiers, as pattern recognition and anomaly detection could be solved as a classification problem. As still a major part of...
-
Asynchronous Charge Carrier Injection in Perovskite Light-Emitting Transistors
PublicationUnbalanced mobility and injection of charge carriers in metal-halide perovskite light-emitting devices pose severe limitations to the efficiency and response time of the electroluminescence. Modulation of gate bias in methylammonium lead iodide light-emitting transistors has proven effective in increasing the brightness of light emission up to MHz frequencies. In this work, a new approach is developed to improve charge carrier...
-
Training of Deep Learning Models Using Synthetic Datasets
PublicationIn order to solve increasingly complex problems, the complexity of Deep Neural Networks also needs to be constantly increased, and therefore training such networks requires more and more data. Unfortunately, obtaining such massive real world training data to optimize neural networks parameters is a challenging and time-consuming task. To solve this problem, we propose an easy-touse and general approach to training deep learning...
-
Evaluation of aspiration problems in L2 English pronunciation employing machine learning
PublicationThe approach proposed in this study includes methods specifically dedicated to the detection of allophonic variation in English. This study aims to find an efficient method for automatic evaluation of aspiration in the case of Polish second-language (L2) English speakers’ pronunciation when whole words are analyzed instead of particular allophones extracted from words. Sample words including aspirated and unaspirated allophones...
-
Comparison of Classification Methods for EEG Signals of Real and Imaginary Motion
PublicationThe classification of EEG signals provides an important element of brain-computer interface (BCI) applications, underlying an efficient interaction between a human and a computer application. The BCI applications can be especially useful for people with disabilities. Numerous experiments aim at recognition of motion intent of left or right hand being useful for locked-in-state or paralyzed subjects in controlling computer applications....
-
How to model ROC curves - a credit scoring perspective
PublicationROC curves, which derive from signal detection theory, are widely used to assess binary classifiers in various domains. The AUROC (area under the ROC curve) ratio or its transformations (the Gini coefficient) belong to the most widely used synthetic measures of the separation power of classification models, such as medical diagnostic tests or credit scoring. Frequently a need arises to model an ROC curve. In the biostatistical...
-
The AFM micrographs of gold nanoparticles on silicon substrate
Open Research DataThe dataset contains the first approach towards AFM topographic imaging of gold nanoparticles synthesized and immobilized on the silicon surface. Measurements were made in the semi-contact mode on the NTEGRA Prima device, manufactured by NT-MDT. Scans were performed with amplitude detection at an operating value of 60% of the free oscillation amplitude....
-
Affective reactions to playing digital games
PublicationThe paper presents a study of emotional states during a gameplay. An experiment of two-player Tetris game is reported, followed by the analysis of the results - self-reported emotional states as well as physiological signals measurements interpretation. The study reveals the diversity of emotional reactions and concludes, that a representative player's emotional model is hard to define. Instead, an adaptive approach to emotion...
-
Improving Effectiveness of SVM Classifier for Large Scale Data
PublicationThe paper presents our approach to SVM implementation in parallel environment. We describe how classification learning and prediction phases were pararellised. We also propose a method for limiting the number of necessary computations during classifier construction. Our method, named one-vs-near, is an extension of typical one-vs-all approach that is used for binary classifiers to work with multiclass problems. We perform experiments...
-
Two Stage SVM and kNN Text Documents Classifier
PublicationThe paper presents an approach to the large scale text documents classification problem in parallel environments. A two stage classifier is proposed, based on a combination of k-nearest neighbors and support vector machines classification methods. The details of the classifier and the parallelisation of classification, learning and prediction phases are described. The classifier makes use of our method named one-vs-near. It is...
-
Layered background modeling for automatic detection of unattended objects in camera images
PublicationAn algorithm for automatic detection of unattended objects in video camera images is presented. First, background subtraction is performed, using an approach based on the codebook method. Results of the detection are then processed by assigning the background pixels to time slots, based on the codeword age. Using this data, moving objects detected during a chosen period may be extracted from the background model. The proposed approach...
-
Sonar Pulse Detection Using Chirp Rate Estimation and CFAR Algorithms
PublicationThis paper presents a new approach to sonar pulse detection. The method uses chirp rate estimators and algorithms for the adaptive threshold, commonly used in radiolocation. The proposed approach allows detection of pulses of unknown parameters, which may be used in passive hydrolocation or jamming detection in underwater communication. Such an analysis is possible thanks to a new kind of imaging, which presents signal energy in...
-
A review of emotion recognition methods based on keystroke dynamics and mouse movements
PublicationThe paper describes the approach based on using standard input devices, such as keyboard and mouse, as sources of data for the recognition of users’ emotional states. A number of systems applying this idea have been presented focusing on three categories of research problems, i.e. collecting and labeling training data, extracting features and training classifiers of emotions. Moreover the advantages and examples of combining standard...
-
Assessment of copper surface coverage with corrosion inhibitor using AFM-based local electrical measurements
PublicationThe paper presents a new method of assessment of metal surface coverage with corrosion inhibitor and thus of inhibitor protective performance. It is based on the atomic force microscopy measurement performed in a contact mode. Apart from topography images the proposed approach allows acquisition of local DC maps and local electrical impedance spectra via application of DC bias voltage or AC perturbation signal between the conductive...
-
Enhanced gas sensing by graphene-silicon Schottky diodes under UV irradiation
PublicationThe effect of ultraviolet (UV) or blue irradiation on graphene/n-doped silicon Schottky junctions toward gas sensing was investigated. Schottky diodes were subjected to oxidizing nitrogen dioxide (NO2, 1–3 ppm) and reducing tetrahydrofuran (THF, 50–200 ppm), showing significantly different responses observed on the currentvoltage (I-V) characteristics, especially under UV light (275 nm). NO2 affected the resistive part of the forward region...