Filters
total: 7955
-
Catalog
- Publications 6490 available results
- Journals 249 available results
- Conferences 107 available results
- People 342 available results
- Projects 19 available results
- Laboratories 3 available results
- e-Learning Courses 318 available results
- Events 10 available results
- Open Research Data 417 available results
displaying 1000 best results Help
Search results for: SURROGATE MODELING , ANTENNA DESIGN , DOMAIN CONFINEMENT , NESTED KRIGING , DEEP NEURAL NETWORKS
-
When Neural Networks Meet Decisional DNA: A Promising New Perspective for Knowledge Representation and Sharing
PublicationABSTRACT In this article, we introduce a novel concept combining neural network technology and Decisional DNA for knowledge representation and sharing. Instead of using traditional machine learning and knowledge discovery methods, this approach explores the way of knowledge extraction through deep learning processes based on a domain’s past decisional events captured by Decisional DNA. We compare our approach with kNN (k-nearest...
-
TR-Based Antenna Design with Forward FD: The Effects of Step Size on the Optimization Performance
PublicationNumerical methods are important tools for design of modern antennas. Trust-region (TR) methods coupled with data-efficient surrogates based on finite differentiation (FD) represent a popular class of antenna design algorithms. However, TR performance is subject to FD setup, which is normally determined a priori based on rules-of-thumb. In this work, the effect of FD perturbations on the performance of TR-based design is evaluated...
-
Globalized Knowledge-Based Simulation-Driven Antenna Miniaturization Using Domain-Confined Surrogates and Dimensionality Reduction
PublicationDesign of contemporary antenna systems encounters multifold challenges, one of which is a limited size. Compact antennas are indispensable for the new fields of application such as inter-net of things or 5G/6G mobile communication. Still, miniaturization generally undermines elec-trical and field performance. When attempted through numerical optimization, it turns into a constrained problem with costly constraints requiring electromagnetic...
-
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...
-
Application of the neural networks for developing new parametrization of the Tersoff potential for carbon
PublicationPenta-graphene (PG) is a 2D carbon allotrope composed of a layer of pentagons having sp2- and sp3-bonded carbon atoms. A study carried out in 2018 has shown that the parameterization of the Tersoff potential proposed in 2005 by Ehrhart and Able (T05 potential) performs better than other potentials available for carbon, being able to reproduce structural and mechanical properties of the PG. In this work, we tried to improve the...
-
Dynamic Bayesian Networks for Symbolic Polyphonic Pitch Modeling
PublicationSymbolic pitch modeling is a way of incorporating knowledge about relations between pitches into the process of an- alyzing musical information or signals. In this paper, we propose a family of probabilistic symbolic polyphonic pitch models, which account for both the “horizontal” and the “vertical” pitch struc- ture. These models are formulated as linear or log-linear interpo- lations of up to fi ve sub-models, each of which is...
-
Multi-objective design of miniaturized impedance transformers by domain segmentation
PublicationFast multi-objective design optimization of compact microstrip impedance transformers is discussed. Our approach exploits approximation models constructed using sampled coarse- mesh EM simulation data in a partitioned design space and response correction techniques for design refinement. Demonstra
-
Design and Optimization of Metamaterial-Based 5G Millimeter Wave Antenna for Gain Enhancement
PublicationIn this brief, a low profile, broadband, high-gain antenna array based on optimized metamaterials (MMs) with dual-beam radiation is reported for 5G millimeters wave (mm-wave) applications. The design is a simple bow tie operating at a 5G band of 28 GHz. It consists of two bow ties with substrate integrated waveguide (SIW)-based power splitter. A broad impedance bandwidth of 26.3−29.8 GHz is obtained by appropriately combining the...
-
Blood Pressure Estimation Based on Blood Flow, ECG and Respiratory Signals Using Recurrent Neural Networks
PublicationThe estimation of systolic and diastolic blood pressure using artificial neural network is considered in the paper. The blood pressure values are estimated using pulse arrival time, and additionally RR intervals of ECG signal together with respiration signal. A single layer recurrent neural network with hyperbolic tangent activation function was used. The average blood pressure estimation error for the data obtained from 21 subjects...
-
Detecting Objects of Various Categories in Optical Remote Sensing Imagery Using Neural Networks
PublicationThe effective detection of objects in remote sensing images is of great research importance, so recent years have seen a significant progress in deep learning techniques in this field. However, despite much valuable research being conducted, many challenges still remain. A lot of research projects focus on detecting objects of a single category (class), while correctly detecting objects of different categories is much harder. The...
-
Buried Object Characterization Using Ground Penetrating Radar Assisted by Data-Driven Surrogate-Models
PublicationThis work addresses artificial-intelligence-based buried object characterization using 3-D full-wave electromagnetic simulations of a ground penetrating radar (GPR). The task is to characterize cylindrical shape, perfectly electric conductor (PEC) object buried in various dispersive soil media, and in different positions. The main contributions of this work are (i) development of a fast and accurate data driven surrogate modeling...
-
Using Convolutional Neural Networks for Corneal Arcus Detection Towards Familial Hypercholesterolemia Screening
PublicationFamilial hypercholesterolemia (FH) is a highly undiagnosed disease. Among FH patients, the onset of premature coronary artery disease is 13 times higher than in the general population. Early diagnosis and treatment is essential to prevent cardiovascular diseases and their complications, and to prolong life. One of the clinical criteria of FH is the occurrence of a corneal arcus (CA) among patients, especially those under 45 years...
-
NETWORKS
Journals -
Simulation-Based Design of Microstrip Linear Antenna Arrays Using Fast Radiation Response Surrogates
PublicationFast yet accurate technique for simulation-based design of linear arrays of microstrip patch antennas is presented. Our technique includes: (i) optimization of the corrected array factor of the antenna array under design for a phase excitation taper resulting in reduced side lobes; (ii) simulation-driven optimization of the array element for element dimensions resulting in matching at and about operational frequency, and (iii)...
-
Bees Detection on Images: Study of Different Color Models for Neural Networks
PublicationThis paper presents an approach to bee detection in video streams using a neural network classifier. We describe the motivation for our research and the methodology of data acquisition. The main contribution to this work is a comparison of different color models used as an input format for a feedforward convolutional architecture applied to bee detection. The detection process has is based on a neural binary classifier that classifies...
-
A new approach to design of weather disruption-tolerant wireless mesh networks
PublicationWireless Mesh Networks, offering transmission rates of 1–10 Gb/s per a millimeter-wave link (utilizing the 71–86 GHz band) seem to be a promising alternative to fiber optic backbone metropolitan area networks because of significantly lower costs of deployment and maintenance. However, despite providing high transmission rates in good weather conditions, high-frequency wireless links are very susceptible to weather disruptions....
-
A Comprehensive Analysis of Deep Neural-Based Cerebral Microbleeds Detection System
PublicationMachine learning-based systems are gaining interest in the field of medicine, mostly in medical imaging and diagnosis. In this paper, we address the problem of automatic cerebral microbleeds (CMB) detection in magnetic resonance images. It is challenging due to difficulty in distinguishing a true CMB from its mimics, however, if successfully solved it would streamline the radiologists work. To deal with this complex three-dimensional...
-
Low-cost multi-objective design of compact microwave structures using domain patching
PublicationA good compromise between size and electrical performance is an important design consideration for compact microwave structures. Comprehensive information about size/performance trade-offs can be obtained through multi-objective optimization. Due to considerable electromagnetic (EM) cross-couplings in highly compressed layouts, the design process has to be conducted at the level of high-fidelity EM analysis which is computationally...
-
Expedited Geometry Scaling of Compact Microwave Passives by Means of Inverse Surrogate Modeling
PublicationIn this paper, the problem of geometry scaling of compact microwave structures is investigated. As opposed to conventional structures (i.e., constructed using uniform transmission lines), re-design of miniaturized circuits (e.g., implemented with artificial transmission lines, ATSs) for different operating frequencies is far from being straightforward due to considerable cross-couplings between the circuit components. Here, we...
-
Fast Fading Influence on the Deep Learning-Based LOS and NLOS Identificationin Wireless Body Area Networks
PublicationIn the article, the fast fading influence on the proposed DL (Deep Learning) approach for LOS (Line-of-Sight) and NLOS (Non-Line-of-Sight) conditions identification in Wireless Body Area Networks is investigated. The research was conducted on the basis of the off-body communication measurements using the developed mobile measurement stand, in an indoor environment for both static and dynamic scenarios. The measurements involved...
-
Expedited EM-Driven Design of Miniaturized Microwave Hybrid Couplers Using Surrogate-Based Optimization
PublicationMiniaturization of microwave hybrid couplers is important for contemporary wireless communication engineering. Using standard computer-aided design methods for development of compact structures is extremely challenging due to a general lack of computationally efficient and accurate simulation models. Poor accuracy of available equivalent circuits results from neglecting parasitic cross-couplings that greatly affect the performance...
-
Miniaturization-Oriented Design of Spline-Parameterized UWB Antenna for In-Door Positioning Applications
PublicationDesign of ultra-wideband antennas for in-door localization applications is a challenging task. It involves development of geometry that maintains appropriate balance between the size and performance. In this work, a topologically-flexible monopole has been generated using a stratified framework which embeds a gradient-based trust-region (TR) optimization algorithm in a meta-loop that gradually increases the structure dimensionality....
-
Gas Detection Using Resistive Gas Sensors And Radial Basis Function Neural Networks
PublicationWe present a use of Radial Basis Function (RBF) neural networks and Fluctuation Enhanced Sensing (FES) method in gas detection system utilizing a prototype resistive WO3 gas sensing layer with gold nanoparticles. We investigated accuracy of gas detection for three different preprocessing methods: no preprocessing, Principal Component Analysis (PCA) and wavelet transformation. Low frequency noise voltage observed in resistive gas...
-
Assessment and design of greener deep eutectic solvents – A multicriteria decision analysis
PublicationDeep eutectic solvents (DES) are often considered as green solvents because of their properties, such as negligible vapor pressure, biodegradability, low toxicity or natural origin of their components. Due to the fact that DES are cheaper than ionic liquids, they have gained many applications in a short period of time. However, claims about their greenness sometimes seem to be exaggerated. Especially, bearing in mind lots of data...
-
Atomistic Surrogate-Based Optimization for Simulation-Driven Design of Computationally Expensive Microwave Circuits with Compact Footprints
PublicationA robust simulation-driven design methodology for computationally expensive microwave circuits with compact footprints has been presented. The general method introduced in this chapter is suitable for a wide class of N-port un-conventional microwave circuits constructed as a deviation from classic design solutions. Conventional electromagnetic (EM) simulation-driven design routines are generally prohibitive when applied to numerically...
-
Circularly Polarized Antenna Array design with the Potential of Gain-Size Trade-off and Omnidirectional Radiation for Millimeter-Wave Small Base Station Applications
PublicationThis paper presents the design and validation of a slot-patch-hybrid circularly polarized antenna array for 28 GHz millimeter (mm) wave (mm-wave) applications. The proposed design has a simple geometry that facilitates the fabrication process, which is otherwise a challenging task due to the sub-mm dimensions of the circuit in the mm-wave band. In the proposed structure, aperture-coupled series slot-fed array is utilized to excite...
-
Neural Network-Based Sequential Global Sensitivity Analysis Algorithm
PublicationPerforming global sensitivity analysis (GSA) can be challenging due to the combined effect of the high computational cost, but it is also essential for engineering decision making. To reduce this cost, surrogate modeling such as neural networks (NNs) are used to replace the expensive simulation model in the GSA process, which introduces the additional challenge of finding the minimum number of training data samples required to...
-
Breast MRI segmentation by deep learning: key gaps and challenges
PublicationBreast MRI segmentation plays a vital role in early diagnosis and treatment planning of breast anomalies. Convolutional neural networks with deep learning have indicated promise in automating this process, but significant gaps and challenges remain to address. This PubMed-based review provides a comprehensive literature overview of the latest deep learning models used for breast segmentation. The article categorizes the literature...
-
Assessment of Therapeutic Progress After Acquired Brain Injury Employing Electroencephalography and Autoencoder Neural Networks
PublicationA method developed for parametrization of EEG signals gathered from participants with acquired brain injuries is shown. Signals were recorded during therapeutic session consisting of a series of computer assisted exercises. Data acquisition was performed in a neurorehabilitation center located in Poland. The presented method may be used for comparing the performance of subjects with acquired brain injuries (ABI) who are involved...
-
Design of compact self-quintuplexing antenna with high-isolation for penta-band applications
PublicationThis article presents a novel compact self-quintuplexing antenna architecture based on a substrate-integrated rectangular cavity (SIRC) for pentaband applications. The proposed self-quintuplexing antenna is constructed by employing an SIRC, one Pi-shaped slot (PSS), one T-shaped slot (TSS), and five 50Ω microstrip feedlines. The PSS and TSS are engraved on the top of the SIRC to create five radiating patches, which are excited...
-
Deep Learning: A Case Study for Image Recognition Using Transfer Learning
PublicationDeep learning (DL) is a rising star of machine learning (ML) and artificial intelligence (AI) domains. Until 2006, many researchers had attempted to build deep neural networks (DNN), but most of them failed. In 2006, it was proven that deep neural networks are one of the most crucial inventions for the 21st century. Nowadays, DNN are being used as a key technology for many different domains: self-driven vehicles, smart cities,...
-
Multi-Objective Design Optimization of Compact Quasi-Isotropic Dielectric Resonator Antenna
PublicationMulti-objective optimization of a quasi-isotropic dielectric resonator antenna (DRA) is presented. Utilization of variable-fidelity electromagnetic (EM) DRA models, response surface approximations, and response correction techniques, allows us to obtain—at a low computational cost—a set of alternative antenna designs representing the best possible trade-offs between three conflicting objectives: antenna size, its reflection response,...
-
Heavy duty vehicle fuel consumption modelling using artificial neural networks
PublicationIn this paper an artificial neural network (ANN) approach to modelling fuel consumption of heavy duty vehicles is presented. The proposed method uses easy accessible data collected via CAN bus of the truck. As a benchmark a conventional method, which is based on polynomial regression model, is used. The fuel consumption is measured in two different tests, performed by using a unique test bench to apply the load to the engine. Firstly,...
-
Modelling of wastewater treatment plant for monitoring and control purposes by state - space wavelet networks
PublicationMost of industrial processes are nonlinear, not stationary, and dynamical with at least few different time scales in their internal dynamics and hardly measured states. A biological wastewater treatment plant falls into this category. The paper considers modelling such processes for monitorning and control purposes by using State - Space Wavelet Neural Networks (SSWN). The modelling method is illustrated based on bioreactors of...
-
Quasi-Global Optimization of Antenna Structures Using Principal Components and Affine Subspace-Spanned Surrogates
PublicationParametric optimization is a mandatory step in the design of contemporary antenna structures. Conceptual development can only provide rough initial designs that have to be further tuned, often extensively. Given the topological complexity of modern antennas, the design closure necessarily involves full-wave electromagnetic (EM) simulations and—in many cases—global search procedures. Both factors make antenna optimization a computationally...
-
Optimal shape design of multi-element trawl-doors using local surrogate models
PublicationTrawl-doors have a large influence on the fuel consumption of fishing vessels. Design and optimiza-tion of trawl-doors using computational models are a key factor in minimizing the fuel consump-tion. This paper presents an optimization algorithm for the shape design of trawl-door shapes using computational fluid dynamic (CFD) models. Accurate CFD models are computationally expensive. Therefore, the direct use of traditional optimization...
-
Zero-Pole Approach in Microwave Passive Circuit Design
PublicationIn this thesis, optimization strategies for design of microwave passive structures including filters, couplers, antenna and impedance transformer and construction of various surroogate models utilized to fasten the design proces have been discussed. Direct and hybrid optimization methodologies including space mapping and multilevel algorithms combined with various surrogate models at different levels of fidelity have been utilized...
-
Towards bees detection on images: study of different color models for neural networks
PublicationThis paper presents an approach to bee detection in videostreams using a neural network classifier. We describe the motivationfor our research and the methodology of data acquisition. The maincontribution to this work is a comparison of different color models usedas an input format for a feedforward convolutional architecture appliedto bee detection. The detection process has is based on a neural...
-
Analysis of circular polarization antenna design trade‐offs using low‐cost EM‐driven multiobjective optimization
PublicationCircular polarization (CP) antennas are vital components of modern communication systems. Their design involves handling several requirements such as low reflection and axial ratio (AR) within the frequency range of interest. Small size is an important criterion for antenna mobility which is normally achieved as a by‐product of performance‐oriented modifications of the structure topology. In this work, multiobjective optimization...
-
Sylwester Kaczmarek dr hab. inż.
PeopleSylwester Kaczmarek received his M.Sc in electronics engineering, Ph.D. and D.Sc. in switching and teletraffic science from the Gdansk University of Technology, Gdansk, Poland, in 1972, 1981 and 1994, respectively. His research interests include: IP QoS and GMPLS and SDN networks, switching, QoS routing, teletraffic, multimedia services and quality of services. Currently, his research is focused on developing and applicability...
-
Data augmentation for improving deep learning in image classification problem
PublicationThese days deep learning is the fastest-growing field in the field of Machine Learning (ML) and Deep Neural Networks (DNN). Among many of DNN structures, the Convolutional Neural Networks (CNN) are currently the main tool used for the image analysis and classification purposes. Although great achievements and perspectives, deep neural networks and accompanying learning algorithms have some relevant challenges to tackle. In this...
-
Design and optimization of a novel miniaturized low-profile circularly polarized wide-slot antenna
PublicationThis paper presents a novel structure of a compact circularly polarized (CP) antenna. CP is obtained using a parasitic quasi-rectangular strip placed coplanar to the feedline. A ground plane perturbation combined with the asymmetric geometry of the coplanar waveguide ground planes is utilized to excite additional CP modes. All antenna dimensions are rigorously optimized to achieve the best possible performance in terms of the impedance...
-
Expedited Design Closure of Antenna Input Characteristics by Trust Region Gradient Search and Principal Component Analysis
PublicationOptimization-based parameter tuning has become an inherent part of contemporary antenna design process. For the sake of reliability, it is typically conducted at the level of full-wave electromagnetic (EM) simulation models. This may incur considerable computational expenses depending on the cost of an individual EM analysis, the number of adjustable variables, the type of task (local, global, single-/multi-objective optimization),...
-
Reduced-Cost Microwave Modeling Using Constrained Domains and Dimensionality Reduction
PublicationDevelopment of modern microwave devices largely exploits full-wave electromagnetic (EM) simulations. Yet, simulation-driven design may be problematic due to the incurred CPU expenses. Addressing the high-cost issues stimulated the development of surrogate modeling methods. Among them, data-driven techniques seem to be the most widespread owing to their flexibility and accessibility. Nonetheless, applicability of approximation-based...
-
Multimode, multiparametric surrogate models for fast design of waveguide components.
PublicationPrzedstawiona została efektywna technika tworzenia wieloparametrycznych mo-deli zastępczych na podstawie wyników symulacji pełnofalowej przy wykorzy-staniu metody Cauchy`ego wielu zmiennych. Metoda pozwala na automatyczny wy-bór rzędu modelu i minimalizuje liczbę koniecznych analiz pełnofalowych. Te-chnika ta jest wykorzystana do tworzenia modeli uogólnionej macierzy rozpro-szenia.
-
Low-cost multi-criterial design optimization of compact microwave passives using constrained surrogates and dimensionality reduction
PublicationDesign of contemporary microwave circuits is a challenging task. Typically, it has to take into account several performance requirements and constraints. The design objectives are often conflicting and their simultaneous improvement may not be possible; instead, compromise solutions are to be sought. Representative examples are miniaturized microwave passives where reduction of the circuit size has a detrimental effect on its electrical...
-
Heavy Duty Vehicle Fuel Consumption Modelling Based on Exploitation Data by Using Artificial Neural Networks
PublicationOne of the ways to improve the fuel economy of heavy duty trucks is to operate the combustion engine in its most efficient operating points. To do that, a mathematical model of the engine is required, which shows the relations between engine speed, torque and fuel consumption in transient states. In this paper, easy accessible exploitation data collected via CAN bus of the heavy duty truck were used to obtain a model of a diesel...
-
The Use of Artificial Neural Networks and Decision Trees to Predict the Degree of Odor Nuisance of Post-Digestion Sludge in the Sewage Treatment Plant Process
PublicationThis paper presents the application of artificial neural networks and decision trees for the prediction of odor properties of post-fermentation sludge from a biological-mechanical wastewater treatment plant. The input parameters were concentrations of popular compounds present in the sludge, such as toluene, p-xylene, and p-cresol, and process parameters including the concentration of volatile fatty acids, pH, and alkalinity in...
-
Neural Network Subgraphs Correlation with Trained Model Accuracy
PublicationNeural Architecture Search (NAS) is a computationally demanding process of finding optimal neural network architecture for a given task. Conceptually, NAS comprises applying a search strategy on a predefined search space accompanied by a performance evaluation method. The design of search space alone is expected to substantially impact NAS efficiency. We consider neural networks as graphs and find a correlation between the presence...
-
Novel approach to modeling spectral-domain optical coherence tomography with Monte Carlo method
PublicationNumerical modeling Optical Coherence Tomography (OCT) systems is needed for optical setup optimization, development of new signal processing methods and assessment of impact of different physical phenomena inside the sample on OCT signal. The Monte Carlo method has been often used for modeling Optical Coherence Tomography, as it is a well established tool for simulating light propagation in scattering media. However, in this method...