• Title/Summary/Keyword: iterative method

Search Result 2,053, Processing Time 0.029 seconds

A Deep Learning Application for Automated Feature Extraction in Transaction-based Machine Learning (트랜잭션 기반 머신러닝에서 특성 추출 자동화를 위한 딥러닝 응용)

  • Woo, Deock-Chae;Moon, Hyun Sil;Kwon, Suhnbeom;Cho, Yoonho
    • Journal of Information Technology Services
    • /
    • v.18 no.2
    • /
    • pp.143-159
    • /
    • 2019
  • Machine learning (ML) is a method of fitting given data to a mathematical model to derive insights or to predict. In the age of big data, where the amount of available data increases exponentially due to the development of information technology and smart devices, ML shows high prediction performance due to pattern detection without bias. The feature engineering that generates the features that can explain the problem to be solved in the ML process has a great influence on the performance and its importance is continuously emphasized. Despite this importance, however, it is still considered a difficult task as it requires a thorough understanding of the domain characteristics as well as an understanding of source data and the iterative procedure. Therefore, we propose methods to apply deep learning for solving the complexity and difficulty of feature extraction and improving the performance of ML model. Unlike other techniques, the most common reason for the superior performance of deep learning techniques in complex unstructured data processing is that it is possible to extract features from the source data itself. In order to apply these advantages to the business problems, we propose deep learning based methods that can automatically extract features from transaction data or directly predict and classify target variables. In particular, we applied techniques that show high performance in existing text processing based on the structural similarity between transaction data and text data. And we also verified the suitability of each method according to the characteristics of transaction data. Through our study, it is possible not only to search for the possibility of automated feature extraction but also to obtain a benchmark model that shows a certain level of performance before performing the feature extraction task by a human. In addition, it is expected that it will be able to provide guidelines for choosing a suitable deep learning model based on the business problem and the data characteristics.

Design of a 10× Zoom Lens with an Expander for an MWIR Camera Using Athermal Material Composition Method (비열화 소재 구성 방법을 이용한 중적외선 카메라용 확장형 10배 줌 렌즈 설계)

  • Ryu, Tae-Sik;Park, Sung-Chan
    • Korean Journal of Optics and Photonics
    • /
    • v.33 no.6
    • /
    • pp.287-294
    • /
    • 2022
  • This study presents a method for designing an athermal middle wavelength infrared (MWIR) zoom lens with the iterative selection of material compositions on an athermal glass map. The optical properties of glass for MWIR are generally very sensitive to temperature, compared with visible glass. To compensate for focus error due to temperature change, the non-athermalized zoom system requires a large amount of movement of a compensator, which results in an unstable zoom system. To solve this problem, the material compositions for an athermal zoom lens have effectively been obtained using the thermal aberration correction process analytically on an athermal glass map. An expander lens is used to enlarge the focal lengths of an original main zoom lens two times. Finally, while this expander is attached to an original athermal zoom system, the final zoom system equipped with this expander doubles the focal length ranges and has stable performance over a specified temperature range.

Deforming the Walking Motion with Geometrical Editing (주 관절 경로의 변형을 통한 걷기 동작 수정)

  • Kim, Meejin;Lee, Sukwon
    • Journal of the Korea Computer Graphics Society
    • /
    • v.27 no.1
    • /
    • pp.1-8
    • /
    • 2021
  • This paper proposes a simple deformation method for editing the trajectory of a walking motion with preserving its style. To this end, our method analyzes the trajectory of the root joint into the graph and deforms it by applying the graph Laplace operator. The trajectory of the root joint is presented as a graph with a vertex defined the position and direction at each time frame on the motion dataThe graph transforms the trajectory into the differential coordinate, and if the constraints are set on the trajectory vertex, the solver iterative approaches to the solution. By modifying the root trajectory, we can continuously vary the walking motion, which reduces the cost of capturing a whole motion that is required. After computes the root trajectory, other joints are copied on the root and post-processed as a final motion. At the end of our paper, we show the application that the character continuously walks in a complex environment while satisfying user constraints.

Iterative Deep Convolutional Grid Warping Network for Joint Depth Upsampling (반복적인 격자 워핑 기법을 이용한 깊이 영상 초해상화 기술)

  • Kim, Dongsin;Yang, Yoonmo;Oh, Byung Tae
    • Journal of Broadcast Engineering
    • /
    • v.25 no.6
    • /
    • pp.965-972
    • /
    • 2020
  • Depth maps have distance information of objects. They play an important role in organizing 3D information. Color and depth images are often simultaneously obtained. However, depth images have lower resolution than color images due to limitation in hardware technology. Therefore, it is useful to upsample depth maps to have the same resolution as color images. In this paper, we propose a novel method to upsample depth map by shifting the pixel position instead of compensating pixel value. This approach moves the position of the pixel around the edge to the center of the edge, and this process is carried out in several steps to restore blurred depth map. The experimental results show that the proposed method improves both quantitative and visual quality compared to the existing methods.

Isogeometric Analysis of FGM Plates in Combination with Higher-order Shear Deformation Theory (등기하해석에 의한 기능경사복합재 판의 역학적 거동 예측)

  • Jeon, Juntai
    • Journal of the Society of Disaster Information
    • /
    • v.16 no.4
    • /
    • pp.832-841
    • /
    • 2020
  • Purpose: This study attempts at analyzing mechanical response of functionally graded material (FGM) plates in bending. An accurate and effective numerical approach based on isogeometric analysis (IGA) combined with higher-order shear deformation plate theory to predict the nonlinear flexural behavior is developed. Method: A higher-order shear deformation theory(HSDT) which accounts for the geometric nonlinearity in the von Karman sense is presented and used to derive the equilibrium and governing equations for FGM plate in bending. The nonlinear equations are solved by the modified Newton-Raphson iterative technique. Result: The volume fraction, plate length-to-thickness ratio and boundary condition have signifiant effects on the nonlinear flexural behavior of FGM plates. Conclusion: The proposed IGA method can be used as an accurate and effective numerical tool for analyzing the mechanical responses of FGM plates in flexure.

Precise System Models using Crystal Penetration Error Compensation for Iterative Image Reconstruction of Preclinical Quad-Head PET

  • Lee, Sooyoung;Bae, Seungbin;Lee, Hakjae;Kim, Kwangdon;Lee, Kisung;Kim, Kyeong-Min;Bae, Jaekeon
    • Journal of the Korean Physical Society
    • /
    • v.73 no.11
    • /
    • pp.1764-1773
    • /
    • 2018
  • A-PET is a quad-head PET scanner developed for use in small-animal imaging. The dimensions of its volumetric field of view (FOV) are $46.1{\times}46.1{\times}46.1mm^3$ and the gap between the detector modules has been minimized in order to provide a highly sensitive system. However, such a small FOV together with the quad-head geometry causes image quality degradation. The main factor related to image degradation for the quad-head PET is the mispositioning of events caused by the penetration effect in the detector. In this paper, we propose a precise method for modelling the system at the high spatial resolution of the A-PET using a LOR (line of response) based ML-EM (maximum likelihood expectation maximization) that allows for penetration effects. The proposed system model provides the detection probability of every possible ray-path via crystal sampling methods. For the ray-path sampling, the sub-LORs are defined by connecting the sampling points of the crystal pair. We incorporate the detection probability of each sub-LOR into the model by calculating the penetration effect. For comparison, we used a standard LOR-based model and a Monte Carlo-based modeling approach, and evaluated the reconstructed images using both the National Electrical Manufacturers Association NU 4-2008 standards and the Geant4 Application for Tomographic Emission simulation toolkit (GATE). An average full width at half maximum (FWHM) at different locations of 1.77 mm and 1.79 mm are obtained using the proposed system model and standard LOR system model, which does not include penetration effects, respectively. The standard deviation of the uniform region in the NEMA image quality phantom is 2.14% for the proposed method and 14.3% for the LOR system model, indicating that the proposed model out-performs the standard LOR-based model.

Application of POD reduced-order algorithm on data-driven modeling of rod bundle

  • Kang, Huilun;Tian, Zhaofei;Chen, Guangliang;Li, Lei;Wang, Tianyu
    • Nuclear Engineering and Technology
    • /
    • v.54 no.1
    • /
    • pp.36-48
    • /
    • 2022
  • As a valid numerical method to obtain a high-resolution result of a flow field, computational fluid dynamics (CFD) have been widely used to study coolant flow and heat transfer characteristics in fuel rod bundles. However, the time-consuming, iterative calculation of Navier-Stokes equations makes CFD unsuitable for the scenarios that require efficient simulation such as sensitivity analysis and uncertainty quantification. To solve this problem, a reduced-order model (ROM) based on proper orthogonal decomposition (POD) and machine learning (ML) is proposed to simulate the flow field efficiently. Firstly, a validated CFD model to output the flow field data set of the rod bundle is established. Secondly, based on the POD method, the modes and corresponding coefficients of the flow field were extracted. Then, an deep feed-forward neural network, due to its efficiency in approximating arbitrary functions and its ability to handle high-dimensional and strong nonlinear problems, is selected to build a model that maps the non-linear relationship between the mode coefficients and the boundary conditions. A trained surrogate model for modes coefficients prediction is obtained after a certain number of training iterations. Finally, the flow field is reconstructed by combining the product of the POD basis and coefficients. Based on the test dataset, an evaluation of the ROM is carried out. The evaluation results show that the proposed POD-ROM accurately describe the flow status of the fluid field in rod bundles with high resolution in only a few milliseconds.

An improved time-domain approach for the spectra-compatible seismic motion generation considering intrinsic non-stationary features

  • Feng Cheng;Jianbo Li;Zhixin Ding;Gao Lin
    • Nuclear Engineering and Technology
    • /
    • v.55 no.3
    • /
    • pp.968-980
    • /
    • 2023
  • The dynamic structural responses are sensitive to the time-frequency content of seismic waves, and seismic input motions in time-history analysis are usually required to be compatible with design response spectra according to nuclear codes. In order to generate spectra-compatible input motions while maintaining the intrinsic non-stationarity of seismic waves, an improved time-domain approach is proposed in this paper. To maintain the nonstationary characteristics of the given seismic waves, a new time-frequency envelope function is constructed using the Hilbert amplitude spectrum. Based on the intrinsic mode functions (IMFs) obtained from given seismic waves through variational mode decomposition, a new corrective time history is constructed to locally modify the given seismic waves. The proposed corrective time history and time-frequency envelope function are unique for each earthquake records as they are extracted from the given seismic waves. In addition, a dimension reduction iterative technique is presented herein to simultaneously superimpose corrective time histories of all the damping ratios at a specific frequency in the time domain according to optimal weights, which are found by the genetic algorithm (GA). Examples are presented to show the capability of the proposed approach in generating spectra-compatible time histories, especially in maintaining the nonstationary characteristics of seismic records. And numerical results reveal that the modified time histories generated by the proposed method can obtain similar dynamic behaviors of AP1000 nuclear power plant with the natural seismic records. Thus, the proposed method can be efficiently used in the design practices.

Analytical study on cable shape and its lateral and vertical sags for earth-anchored suspension bridges with spatial cables

  • Gen-min Tian;Wen-ming Zhang;Jia-qi Chang;Zhao Liu
    • Structural Engineering and Mechanics
    • /
    • v.87 no.3
    • /
    • pp.255-272
    • /
    • 2023
  • Spatial cable systems can provide more transverse stiffness and torsional stiffness without sacrificing the vertical bearing capacity compared with conventional vertical cable systems, which is quite lucrative for long-span earth-anchored suspension bridges' development. Higher economy highlights the importance of refined form-finding analysis. Meanwhile, the internal connection between the lateral and vertical sags has not yet been specified. Given this, an analytic algorithm of form-finding for the earth-anchored suspension bridge with spatial cables is proposed in this paper. Through the geometric compatibility condition and mechanical equilibrium condition, the expressions for cable segment, the recurrence relationship between catenary parameters and control equations of spatial cable are established. Additionally, the nonlinear general reduced gradient method is introduced into fast and high-precision numerical analysis. Furthermore, the analytic expression of the lateral and vertical sags is deduced and discussed. This is very significant for the space design above the bridge deck and the optimization of the sag-to-span ratio in the preliminary design stage of the bridge. Finally, the proposed method is verified with the aid of two examples, one being an operational self-anchored suspension bridge (with spatial cables and a 260 m main span), and the other being an earth-anchored suspension bridge under design (with spatial cables and a 500 m main span). The necessity of an iterative calculation for hanger tensions on earth-anchored suspension bridges is confirmed. It is further concluded that the main cable and their connected hangers are in very close inclined planes.

The Study of Reducing Radiation Exposure Dose and Comparing SUV According to Applied IRIS (Iterative Reconstruction in Image Space) for PET/CT (PET/CT 검사 시 IRIS (Iterative Reconstruction in Image Space) 적용에 따른 CT 피폭선량 감소와 PET SUV 비교 연구)

  • Do, Yong Ho;Song, Ho Jun;Lee, Hyung Jin;Lee, Hong Jae;Kim, Jin Eui
    • The Korean Journal of Nuclear Medicine Technology
    • /
    • v.16 no.2
    • /
    • pp.29-34
    • /
    • 2012
  • Purpose : Presently, hardwares and softwares for reducing radiation exposure are continually developed for PET/CT examination. Purpose of this study is to evaluate effectiveness of reducing radiation exposure dose of CT and SUV changes of PET when applied each kernel to ACCT (Attenuation Correction Computed Tomography) according to adopted IRIS (Iterative Reconstruction in Image Space) software. Materials and Methods : Biograph mCT (Siemens, Germany) was used as a PET/CT scanner. Using AAPM CT performance phantom, from standard (120 kVp, 100 mAs), 7 scans were conducted by reducing 15 mAs each. After image reconstruction by FBP (Filtered Back Projection) and IRIS, noise and spatial resolution were evaluated. The same method was applied to anthropomorphic chest phantom and acquired images were compared. NEMA IEC body phantom was used for SUV evaluation. Injected dose rate for hot sphere (hot) and background cylinder (BKG) were 1:8. CT dose condition (120 kVp, 50 mAs) was the same for each scan and PET scan durations were 1, 2, 3 and 4min. After scanning, each kernel of IRIS was applied to ACCT. And PET images were reconstructed by ACCT adopted IRIS for comparing SUV changes. Results : AAPM phantom test for noise evaluation, SD for FBP 100 mAs, IRIS 55 mAs were 8.8 and 8.9. FBP 85 mAs, IRIS 40 mAs were 9.5 and 9.7. FBP 70 mAs, IRIS 25 mAs were 11.9 and 11.1. Above mAs condition for FBP and IRIS, SD showed similar values. And for spatial resolution test, there was no significant difference. For chest phantom test, when applied the same mAs and kernel to both of FBP and IRIS, every applied kernels showed reduced noise. Lower mAs and higher kernel value showed higher noise reduction. There was no considerable difference only except for I70 very sharp kernel for SUV comparison using NEMA IEC body phantom. Conclusion : In this study, low mAs (55 mAs) applied IRIS and standard mAs (100 mAs) applied FBP showed similar noise. And only except for I70 kernel, there was no significant SUV changes. It is possible to reduce needless radiation exposure and acquire better image quality than FBP's through applying appropriate kernel of IRIS to PET/CT.

  • PDF