Entity

Time filter

Source Type


Kossakowski P.,Warsaw University of Life Sciences | Bilski P.,Warsaw University of Life Sciences | Bilski P.,Institute of Radioelectronics
Proceedings of the 2015 IEEE 8th International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2015 | Year: 2015

In this paper the application of Self-Organizing Maps to analyze the stock market data is presented. The impact of this type of the neural network on the performance of investment strategies is discussed. The considered characteristics include the size of the network based on the average number of learning patterns per neuron, conscience mechanism, method of weights update and learning coefficient. Performance of each network configuration is verified against the simple investment strategy, which on the basis of average Rate of Return (RoR) generates 'buy' and 'sell' signals. Results show that SOMs with the conscience mechanism performs better than the counterpart without it. © 2015 IEEE.


Bilski P.,Institute of Radioelectronics
Geotechnical Engineering for Infrastructure and Development - Proceedings of the XVI European Conference on Soil Mechanics and Geotechnical Engineering, ECSMGE 2015 | Year: 2015

The paper presents the application of the data clustering methods for the automated generation of the 3D soil profile. The software was created for this purpose, acquiring multiple measurement data sets from separate probing locations in the selected site and performing data analysis. The latter was conducted based on multiple in-situ tests involving CPT and DMT investigations. This way it was possible to generate profiles using only one type of the probe or to combine data from multiple probes to obtain the profile with greater accuracy. Popular unsupervised learning approaches, i.e. graph clustering and nearest neighbour were used for data analysis and their performance compared. The conducted experiments consisted in generating soil categories based on multiple sets, presenting them in the graphical, three-dimensional form. Parameters of clustering methods (such as the threshold separating different soil categories or the number of desired groups) and their impact on the obtained profiles were verified. The approach was tested in selected sites within the Warsaw University of Life Sciences campus. The obtained profiles were also compared to the ones generated based on the boreholes, which allowed for confirming their usefulness. The resolution of the approach (the number of different categories obtainable) was also tested. © The authors and ICE Publishing: All rights reserved, 2015.


Coquillat D.,CNRS Charles Coulomb Laboratory | Marczewski J.,Polish Institute of Electron Technology | Kopyt P.,Institute of Radioelectronics | Dyakonova N.,CNRS Charles Coulomb Laboratory | And 3 more authors.
Optics Express | Year: 2016

Phenomena of the radiation coupling to the field effect transistors based terahertz (THz) detectors are studied. We show that in the case of planar metal antennas a significant portion of incoming radiation, instead of being coupled to the transistors, is coupled to an antenna substrate leading to responsivity losses and/or cross-talk effects in the field effect based THz detector arrays. Experimental and theoretical investigations of the responsivity versus substrate thickness are performed. They clearly show how to minimize the losses by the detector/ array substrate thinning. In conclusion simple quantitative rules of losses minimization by choosing a proper substrate thickness of field effect transistor THz detectors are presented for common materials (Si, GaAs, InP, GaN) used in semiconductor technologies. ©2016 Optical Society of America.


But D.B.,CNRS Charles Coulomb Laboratory | Coquillat D.,CNRS Charles Coulomb Laboratory | Dyakonova N.,CNRS Charles Coulomb Laboratory | Teppe F.,CNRS Charles Coulomb Laboratory | And 4 more authors.
2016 21st International Conference on Microwave, Radar and Wireless Communications, MIKON 2016 | Year: 2016

Experimental and theoretical investigations of the influence of substrate thickness on the performance of Si field effect transistor terahertz detectors are presented.. We show that in the case of planar metal antennas a significant portion of incoming radiation, instead of being coupled to the transistors, may be coupled to the substrate leading to important responsivity losses..The ways of avoiding these losses are discussed. © 2016 IEEE.


Coquillat D.,CNRS Charles Coulomb Laboratory | Marczewski J.,Polish Institute of Electron Technology | Kopyt P.,Institute of Radioelectronics | Dyakonova N.,CNRS Charles Coulomb Laboratory | And 7 more authors.
IRMMW-THz 2015 - 40th International Conference on Infrared, Millimeter, and Terahertz Waves | Year: 2015

Phenomena of the radiation coupling to the field effect transistors based terahertz detectors is studied. We show that in the case of flat metallic antennas important part of radiation, instead coupling to the transistors, is coupled to the substrate leading to losses. Experimental and theoretical investigations of the responsivity versus substrate thickness are performed. They clearly show how to minimize the losses by the detector substrate thinning. © 2015 IEEE.

Discover hidden collaborations