• Proceedings of the Korean Ceranic Society Conference
• /
• 2003.04a
• /
• pp.35.1-35
• /
• 2003

• Electrical & Electronic Materials
• /
• v.7 no.3
• /
• pp.213-219
• /
• 1994

Cu compensated Si doped GaAs (GaAs :Si:Cu has been chosen as the switch material. The GaAs material has been characterized by DLTS(Deep Level Transient Spectroscopy) technique and the obtained data were used in the computer simulation. Simulation studies are performed on several GaAs switch systems, composed of different densities of Cu, to investigate the influence of deep traps in the switch systems. The computed results demonstrates important aspect of the switch, the existence of two stable states and fast optical quenching. An important parameter optimum Cu density for the switch are also determined.

• Smart Structures and Systems
• /
• v.31 no.4
• /
• pp.383-392
• /
• 2023

Post-earthquake building condition assessment is crucial for subsequent rescue and remediation and can be automated by emerging computer vision and deep learning technologies. This study is based on an endeavour for the 2nd International Competition of Structural Health Monitoring (IC-SHM 2021). The task package includes five image segmentation objectives - defects (crack/spall/rebar exposure), structural component, and damage state. The structural component and damage state tasks are identified as the priority that can form actionable decisions. A multi-task Convolutional Neural Network (CNN) is proposed to conduct the two major tasks simultaneously. The rest 3 sub-tasks (spall/crack/rebar exposure) were incorporated as auxiliary tasks. By synchronously learning defect information (spall/crack/rebar exposure), the multi-task CNN model outperforms the counterpart single-task models in recognizing structural components and estimating damage states. Particularly, the pixel-level damage state estimation witnesses a mIoU (mean intersection over union) improvement from 0.5855 to 0.6374. For the defect detection tasks, rebar exposure is omitted due to the extremely biased sample distribution. The segmentations of crack and spall are automated by single-task U-Net but with extra efforts to resample the provided data. The segmentation of small objects (spall and crack) benefits from the resampling method, with a substantial IoU increment of nearly 10%.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2016.02a
• /
• pp.246-246
• /
• 2016

Tungsten-nitrogen (W-N) co-doping has been known to enhance the photocatalytic activity of anatase titania nanoparticles by utilizing visible light. The doping effects are, however, largely dependent on calcination or annealing conditions, and thus, the massive production of quality-controlled photocatalysts still remains a challenge. Using density functional theory (DFT) thermodynamics and time-dependent DFT (TDDFT) computations, we investigate the atomic structures of N doping and W-N co-doping in anatase titania, as well as the effect of the thermal processing conditions. We find that W and N dopants predominantly constitute two complex structures: an N interstitial site near a Ti vacancy in the triple charge state and the simultaneous substitutions of Ti by W and the nearest O by N. The latter case induces highly localized shallow in-gap levels near the conduction band minimum (CBM) and the valence band maximum (VBM), whereas the defect complex yielded deep levels (1.9 eV above the VBM). Electronic structures suggest that substitutions of Ti by W and the nearest O by N improves the photocatalytic activity of anatase by band gap narrowing, while defective structure degrades the activity by an in-gap state-assisted electron-hole recombination, which explains the experimentally observed deep level-related photon absorption. Through the real-time propagation of TDDFT (rtp-TDDFT), we demonstrate that the presence of defective structure attracts excited electrons from the conduction band to a localized in-gap state within a much shorter time than the flat band lifetime of titania. Based on these results, we suggest that calcination under N-rich and O-poor conditions is desirable to eliminate the deep-level states to improve photocatalysis.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2013.02a
• /
• pp.420-420
• /
• 2013

We investigated oxygen plasma effect on defect states near the interface of AlGaN/GaN High Electron Mobility Transistor (HEMT) structure grown on a silicon substrate. After the plasma treatment, electrical properties were evaluated using a frequency dependant Capacitance-Voltage (C-V) and a temperature dependant C-V measurements, and a deep level transient spectroscopy (DLTS) method to study the change of defect densities. In the depth profile resulted from the temperature dependant C-V, a sudden decrease in the carrier concentration for two-dimensional electron gas (2DEG) nearby 250 K was observed. In C-V measurement, the interface states were improved in case of the oxygen-plasma treated samples, whereas the interface was degraded in case of the nitrogen-plasma treated sample. In the DLTS measurement, it was observed the two kinds of defects well known in AlGaN/GaN structure grown on sapphire substrate, which have the activation energies of 0.15 eV, 0.25 eV below the conduction band. We speculate that this defect state in AlGaN/GaN on the silicon substrate is caused from the decrease in 2DEG's carrier concentrations. We compared the various DLTS signals with filling pulse times to identify the characteristics of the newly found defect. In the filling pulse time range under the 80 us, the activation energies changed as the potential barrier model. On the other hand, in the filling pulse time range above the 80 us, the activation energies changed as the extended potential model. Therefore, we suggest that the found defect in the AlGaN/GaN/Si structure could be the extended defect related with AlGa/N/GaN interface states.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.38 no.10
• /
• pp.678-686
• /
• 2001

Origins for the transconductance dispersion and the gate leakage current in a GaAs metal semiconductor field effect transistor were found using capacitance deep-level transient spectroscopy (DLTS) measurements. In DLTS spectra, we observed two surface states with thermal activation energies of 0.65 $\times$ 0.07 eV and 0.88 $\times$ 0.04 eV and an electron trap EL2 with thermal activation energy of 0.84 $\times$ 0.01 eV. Transconductance was decreased in the frequency range of 5.5 Hz ~ 300 Hz. The transition frequency shifted to higher frequencies with the increase of temperature and the activation energy for the change of the transition frequency was determined to be 0.66 $\times$ 0.02 eV. From the measurements of the gate leakage current as a function of the device temperature, the forward and reverse currents are coincident with each other below gate voltages lower than 0.15 V, namely Ohmic behavior between gate and source/drain electrodes. The activation energy for the conductance of electrons on the surface of MESFET was 0.63 $\times$ 0.01 eV. Comparing activation energies obtained by different measurements, we found surface states H1 caused the transconductance dispersion and the fate leakage current.

• Journal of Korea Society of Industrial Information Systems
• /
• v.26 no.1
• /
• pp.1-10
• /
• 2021

This study proposed a Long-Short Term Memory network to consider changes in emotion over time, and applied an attention mechanism to give weights to the emotion states that appear at specific moments. We used 32 channel EEG data from DEAP database. A 2-level classification (Low and High) experiment and a 3-level classification experiment (Low, Middle, and High) were performed on Valence and Arousal emotion model. As a result, accuracy of the 2-level classification experiment was 90.1% for Valence and 88.1% for Arousal. The accuracy of 3-level classification was 83.5% for Valence and 82.5% for Arousal.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2009.06a
• /
• pp.327-327
• /
• 2009

ZnO with a large band gap (~3.37 eV) and exciton binding energy (~60 meV), is suitable for optoelectronic applications such as ultraviolet (UV) light emitting diodes (LEDs) and detectors. However, the ZnO-based p-n homojunction is not readily available because it is difficult to fabricate reproducible p-type ZnO with high hall concentration and mobility. In order to solve this problem, there have been numerous attempts to develop p-n heterojunction LEDs with ZnO as the n-type layer. The n-ZnO/p-GaN heterostructure is a good candidate for ZnO-based heterojunction LEDs because of their similar physical properties and the reproducible availability of p-type GaN. Especially, the reduced lattice mismatch (~1.8 %) and similar crystal structure result in the advantage of acquiring high performance LED devices. In particular, a number of ZnO films show UV band-edge emission with visible deep-level emission, which is originated from point defects such as oxygen vacancy, oxygen interstitial, zinc interstitial[1]. Thus, defect-related peak positions can be controlled by variation of growth or annealing conditions. In this work, the undoped ZnO film was grown on the p-GaN:Mg film using RF magnetron sputtering method. The undoped ZnO/p-GaN:Mg heterojunctions were annealed in a horizontal tube furnace. The annealing process was performed at $800^{\circ}C$ during 30 to 90 min in air ambient to observe the variation of the defect states in the ZnO film. Photoluminescence measurements were performed in order to confirm the deep-level position of the ZnO film. As a result, the deep-level emission showed orange-red color in the as-deposited film, while the defect-related peak positions of annealed films were shifted to greenish side as increasing annealing time. Furthermore, the electrical resistivity of the ZnO film was decreased after annealing process. The I-V characteristic of the LEDs showed nonlinear and rectifying behavior. The room-temperature electroluminescence (EL) was observed under forward bias. The EL showed a weak white and strong yellowish emission colors (~575 nm) in the undoped ZnO/p-GaN:Mg heterojunctions before and after annealing process, respectively.

• International journal of advanced smart convergence
• /
• v.11 no.1
• /
• pp.19-27
• /
• 2022

Across the world, 'housing' comprises a significant portion of wealth and assets. For this reason, fluctuations in real estate prices are highly sensitive issues to individual households. In Korea, housing prices have steadily increased over the years, and thus many Koreans view the real estate market as an effective channel for their investments. However, if one purchases a real estate property for the purpose of investing, then there are several risks involved when prices begin to fluctuate. The purpose of this study is to design a real estate price 'return rate' prediction model to help mitigate the risks involved with real estate investments and promote reasonable real estate purchases. Various approaches are explored to develop a model capable of predicting real estate prices based on an understanding of the immovability of the real estate market. This study employs the LSTM method, which is based on artificial intelligence and deep learning, to predict real estate prices and validate the model. LSTM networks are based on recurrent neural networks (RNN) but add cell states (which act as a type of conveyer belt) to the hidden states. LSTM networks are able to obtain cell states and hidden states in a recursive manner. Data on the actual trading prices of apartments in autonomous districts between January 2006 and December 2019 are collected from the Actual Trading Price Disclosure System of the Ministry of Land, Infrastructure and Transport (MOLIT). Additionally, basic data on apartments and commercial buildings are collected from the Public Data Portal and Seoul Metropolitan Government's data portal. The collected actual trading price data are scaled to monthly average trading amounts, and each data entry is pre-processed according to address to produce 168 data entries. An LSTM model for return rate prediction is prepared based on a time series dataset where the training period is set as April 2015~August 2017 (29 months), the validation period is set as September 2017~September 2018 (13 months), and the test period is set as December 2018~December 2019 (13 months). The results of the return rate prediction study are as follows. First, the model achieved a prediction similarity level of almost 76%. After collecting time series data and preparing the final prediction model, it was confirmed that 76% of models could be achieved. All in all, the results demonstrate the reliability of the LSTM-based model for return rate prediction.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.17 no.1
• /
• pp.58-67
• /
• 1992

Optical absorption and photoluminescences(PL) of MgGa2Se4 and MgGa2Se4 : Co2+ single crustals were guown by the Bridgman method have been investigated in the visible and near-in frared regions. The optical absorption spectrum showed three absorption peak at 760 nm(13158nm, -1, 1.63eV), 1640nm(6097cm-1, 0.75eV).and 2500nm(4000cm-1,0.49eV) which are assigned the electronic transitions between the ground state and excited states of Co2+ ions with Td sym-metry in MgGa2Se4 host lattice. In PL spectrum the visible emission bands as well as the infrared emission band in these single cuystals are obserned. The visible emission bands are explained due to the radiative transitions of electrons from quasi continusly distributed tarps below the bottom of the conduction band to acceptor levels above the top of the valence band in the proposed energy level scheme. At the same time, it is considered that the infrated emission bands are attributed to electron transitions from the deep levels to the acceptor levels. The mechanism of the optical transition os well explained in terms of the energy diagram of MgGa2Se4.