• Title/Summary/Keyword: Unit Vector

Search Result 446, Processing Time 0.02 seconds

Large-scale 3D fast Fourier transform computation on a GPU

  • Jaehong Lee;Duksu Kim
    • ETRI Journal
    • /
    • v.45 no.6
    • /
    • pp.1035-1045
    • /
    • 2023
  • We propose a novel graphics processing unit (GPU) algorithm that can handle a large-scale 3D fast Fourier transform (i.e., 3D-FFT) problem whose data size is larger than the GPU's memory. A 1D FFT-based 3D-FFT computational approach is used to solve the limited device memory issue. Moreover, to reduce the communication overhead between the CPU and GPU, we propose a 3D data-transposition method that converts the target 1D vector into a contiguous memory layout and improves data transfer efficiency. The transposed data are communicated between the host and device memories efficiently through the pinned buffer and multiple streams. We apply our method to various large-scale benchmarks and compare its performance with the state-of-the-art multicore CPU FFT library (i.e., fastest Fourier transform in the West [FFTW]) and a prior GPU-based 3D-FFT algorithm. Our method achieves a higher performance (up to 2.89 times) than FFTW; it yields more performance gaps as the data size increases. The performance of the prior GPU algorithm decreases considerably in massive-scale problems, whereas our method's performance is stable.

Forecasting realized volatility using data normalization and recurrent neural network

  • Yoonjoo Lee;Dong Wan Shin;Ji Eun Choi
    • Communications for Statistical Applications and Methods
    • /
    • v.31 no.1
    • /
    • pp.105-127
    • /
    • 2024
  • We propose recurrent neural network (RNN) methods for forecasting realized volatility (RV). The data are RVs of ten major stock price indices, four from the US, and six from the EU. Forecasts are made for relative ratio of adjacent RVs instead of the RV itself in order to avoid the out-of-scale issue. Forecasts of RV ratios distribution are first constructed from which those of RVs are computed which are shown to be better than forecasts constructed directly from RV. The apparent asymmetry of RV ratio is addressed by the Piecewise Min-max (PM) normalization. The serial dependence of the ratio data renders us to consider two architectures, long short-term memory (LSTM) and gated recurrent unit (GRU). The hyperparameters of LSTM and GRU are tuned by the nested cross validation. The RNN forecast with the PM normalization and ratio transformation is shown to outperform other forecasts by other RNN models and by benchmarking models of the AR model, the support vector machine (SVM), the deep neural network (DNN), and the convolutional neural network (CNN).

Development of Line Density Index for the Quantification of Oceanic Thermal Fronts (해양의 수온전선 정량화를 위한 선밀도 지수 개발)

  • Cho, Hyun-Woo;Kim, Kye-Hyun
    • Journal of the Korean Association of Geographic Information Studies
    • /
    • v.9 no.2
    • /
    • pp.227-238
    • /
    • 2006
  • Line density index(LDI) was developed to quantify a densely isothermal line rate as standard index in the ocean environment. Theoretical background on the LDI development process restricting index range 0 to 100 was described. And validation test was done for the LDI application condition that total line length is not greater than 1/10 of unit area. NOAA SST(Sea Surface Temperature) data were used for the experimental application of LDI in the South Sea of Korea. Using GIS, $0.1^{\circ}C$ isothermal lines were linearized as vector data form SST raster data, and unit area were built as polygon data. For the LDI calculation, spatial overlapping(line in polygon) was implemented. To analyze the effect of unit area size for the LDI distribution, two cases of unit area size were designed and descriptive statistics was calculated including performing normality test. The results showed no change of LDI's essential characteristics such as mean and normality except for the range of value, variance and standard deviation. Accordingly, it was found that complex structure of thermal front and even smaller scale of front width than unit area size could influence on the LDI distribution. Also, correlation analysis performed between LDI and difference of temperature(${\Delta}T^{\circ}C$), and horizontal thermal gradient(${\Delta}T^{\circ}C/km$) on the front was obtained from linear regression model. This obtained value was compared with the results from previous researches. Newly developed LDI can be used to compare the thermal front regions changing spatio-temporally in the ocean environment using absolute index value. It is considered to be significant to analyze the relationship between thermal front and marine environment or front and marine organisms in a quantitative approach described in this study.

  • PDF

Constitutive Expression of Arylsulfatase from Pseudoalteromonas carageenovora in E. coli and Its Application to Preparation of Agarose (E. coli에서 Pseudoalteromonas carageenovora 유래 Arylsulfatase의 구성적 발현과 Agarose 제조에의 응용)

  • Kim, Mi-Jin;Jang, Yhon-Hwa;Sung, Moon-Hee;Kim, Yeon-Hee;Nam, Soo-Wan
    • Microbiology and Biotechnology Letters
    • /
    • v.35 no.1
    • /
    • pp.11-16
    • /
    • 2007
  • The arylsulfatase gene (astA, 984 bp ORF) from Pseudoalteromonas carrageenovora genome was amplified by PCR and subcloned into the pHCE-IA vector, in which the hyper consitutive expression (HCE) promoter from the D-amino acid aminotransferase (D-AAT) gene of Geobacillus toevii was employed. The transformant cell, Escherichia coli BL21 (DE3)/pHCE-AST, on LB agar plate containig 4-methylumbelliferyl sulfate, showed an intense fluorescence at 360 nm, indicating that 4-methylumbelliferone was liberated by desulfatate activity. When BL21 (DE3)/pHCE-AST was grown on LB media containing 0.4% glucose or 0.4% glycerol, the arylsulfatase activity was higher at glycerol rather than at glucose. On 2% glycerol medium, the arylsulfatase activity reached 15.0 unit/ml, which was 2.6-fold higher expression level than that with 1% glycerol. The DNA ladder in agarose prepared from agar by this recombinant enzyme revealed similar resolution and migration patterns with a commercial agarose. This results suggests that arylsulfatase overexpressed in E. coli could be applicable to the economic production of electrophoretic-grade agarose.

Effects of Ribosomal Protein L39-L on the Drug Resistance Mechanisms of Lung Cancer A549 Cells

  • Liu, Hong-Sheng;Tan, Wen-Bin;Yang, Ning;Yang, Yuan-Yuan;Cheng, Peng;Liu, Li-Juan;Wang, Wei-Jie;Zhu, Chang-Liang
    • Asian Pacific Journal of Cancer Prevention
    • /
    • v.15 no.7
    • /
    • pp.3093-3097
    • /
    • 2014
  • Background: Cancer is a major threat to the public health whether in developed or in developing countries. As the most common primary malignant tumor, the morbidity and mortality rate of lung cancer continues to rise in recent ten years worldwide. Chemotherapy is one of the main methods in the treatment of lung cancer, but this is hampered by chemotherapy drug resistance, especially MDR. As a component of the 60S large ribosomal subunit, ribosomal protein L39-L gene was reported to be expressed specifically in the human testis and human cancer samples of various tissue origins. Materials and Methods: Total RNA of cultured drug-resistant and susceptible A549 cells was isolated, and real time quantitative RT-PCR were used to indicate the transcribe difference between amycin resistant and susceptible strain of A549 cells. Viability assay were used to show the amycin resistance difference in RPL39-L transfected A549 cell line than control vector and null-transfected A549 cell line. Results: The ribosomal protein L39-L transcription level was 8.2 times higher in drug-resistant human lung cancer A549 cell line than in susceptible A549 cell line by quantitative RT-PCR analysis. The ribosomal protein L39-L transfected cells showed enhanced drug resistance compared to plasmid vector-transfected or null-transfected cells as determined by methyl tritiated thymidine (3H-TdR) incorporation. Conclusions and Implications for Practice: The ribosomal protein L39-L gene may have effects on the drug resistance mechanism of lung cancer A549 cells.

Cell Surface Display of Cycloinulooligosaccharide Fructanotransferase Gene in Saccharomyces cerevisiae (Saccharomyces cerevisiae에서 Cycloinulooligosaccharide Fructanotransferase 유전자의 표층 발현)

  • Kim, Hyun-Jin;Lee, Jae-Hyung;Kim, Hyun-Chul;Kim, Yeon-Hee;Kwon, Hyun-Ju;Nam, Soo-Wan
    • Journal of Life Science
    • /
    • v.17 no.2 s.82
    • /
    • pp.241-247
    • /
    • 2007
  • The cycloinulooligosaccharide fructanotransferase (CFTase) gene (cft) from Paenibacillus macerans was subcloned into the surface display vector, pCTcon (GAL1 promoter). The constructed plasmid, pCTECFTN (9.0 kb) was introduced to S. cerevisiae EBY100 cell and then east transformants were selected on the synthetic defined medium lacking uracil and on the inulin containing medium. The surface display of CFTase was confirmed by immunofluorescence microscopy and its enzymatic ability to form cycloinulooligosaccharides(cyclofructans, CFs) from inulin. The total activity of the CFTase was reached about 5.52 unit/1 by cultivation of yeast transformant on YPDG medium. The optimized conditions determined were as follows; pH, 8.0; temperature, $50^{\circ}C$ ; substrate concentration, 5%; inulin source, Jerusalem artichoke. By the reaction with inulin, CFs consisting of cycloinulohexaose (CF6), cycloinuloheptaose (CF7), and cycloinulooctaose (CF8) were produced and CF6 was the major product.

Five Newly Collected Turnip Mosaic Virus (TuMV) Isolates from Jeju Island, Korea are Closely Related to Previously Reported Korean TuMV Isolates but Show Distinctive Symptom Development

  • Hu, Wen-Xing;Kim, Byoung-Jo;Kwak, Younghwan;Seo, Eun-Young;Kim, Jung-Kyu;Han, Jae-Yeong;Kim, Ik-Hyun;Lim, Yong Pyo;Cho, In-Sook;Domier, Leslie L;Hammond, John;Lim, Hyoun-Sub
    • The Plant Pathology Journal
    • /
    • v.35 no.4
    • /
    • pp.381-388
    • /
    • 2019
  • For several years, temperatures in the Korean peninsula have gradually increased due to climate change, resulting in a changing environment for growth of crops and vegetables. An associated consequence is that emerging species of insect vector have caused increased viral transmission. In Jeju Island, Korea, occurrences of viral disease have increased. Here, we report characterization of five newly collected turnip mosaic virus (TuMV) isolates named KBJ1, KBJ2, KBJ3, KBJ4 and KBJ5 from a survey on Jeju Island in 2017. Full-length cDNAs of each isolate were cloned into the pJY vector downstream of cauliflower mosaic virus 35S and bacteriophage T7 RNA polymerase promoters. Their fulllength sequences share 98.9-99.9% nucleotide sequence identity and were most closely related to previously reported Korean TuMV isolates. All isolates belonged to the BR group and infected both Chinese cabbage and radish. Four isolates induced very mild symptoms in Nicotiana benthamiana but KBJ5 induced a hypersensitive response. Symptom differences may result from three amino acid differences uniquely present in KBJ5; Gly(382)Asp, Ile(891)Val, and Lys(2522)Glu in P1, P3, and NIb, respectively.

Performance Improvement by a Virtual Documents Technique in Text Categorization (문서분류에서 가상문서기법을 이용한 성능 향상)

  • Lee, Kyung-Soon;An, Dong-Un
    • The KIPS Transactions:PartB
    • /
    • v.11B no.4
    • /
    • pp.501-508
    • /
    • 2004
  • This paper proposes a virtual relevant document technique in the teaming phase for text categorization. The method uses a simple transformation of relevant documents, i.e. making virtual documents by combining document pairs in the training set. The virtual document produced by this method has the enriched term vector space, with greater weights for the terms that co-occur in two relevant documents. The experimental results showed a significant improvement over the baseline, which proves the usefulness of the proposed method: 71% improvement on TREC-11 filtering test collection and 11% improvement on Routers-21578 test set for the topics with less than 100 relevant documents in the micro average F1. The result analysis indicates that the addition of virtual relevant documents contributes to the steady improvement of the performance.

Generation of an Infectious Clone of a New Korean Isolate of Apple chlorotic leaf spot virus Driven by Dual 35S and T7 Promoters in a Versatile Binary Vector

  • Kim, Ik-Hyun;Han, Jae-Yeong;Cho, In-Sook;Ju, HyeKyoung;Moon, Jae Sun;Seo, Eun-Young;Kim, Hong Gi;Hammond, John;Lim, Hyoun-Sub
    • The Plant Pathology Journal
    • /
    • v.33 no.6
    • /
    • pp.608-613
    • /
    • 2017
  • The full-length sequence of a new isolate of Apple chlorotic leaf spot virus (ACLSV) from Korea was divergent, but most closely related to the Japanese isolate A4, at 84% nucleotide identity. The full-length cDNA of the Korean isolate of ACLSV was cloned into a binary vector downstream of the bacteriophage T7 RNA promoter and the Cauliflower mosaic virus 35S promoter. Chenopodium quinoa was successfully infected using in vitro transcripts synthesized using the T7 promoter, detected at 20 days post inoculation (dpi), but did not produce obvious symptoms. Nicotiana occidentalis and C. quinoa were inoculated through agroinfiltration. At 32 dpi the infection rate was evaluated; no C. quinoa plants were infected by agroinfiltration, but infection of N. occidentalis was obtained.

Hand Gesture Interface Using Mobile Camera Devices (모바일 카메라 기기를 이용한 손 제스처 인터페이스)

  • Lee, Chan-Su;Chun, Sung-Yong;Sohn, Myoung-Gyu;Lee, Sang-Heon
    • Journal of KIISE:Computing Practices and Letters
    • /
    • v.16 no.5
    • /
    • pp.621-625
    • /
    • 2010
  • This paper presents a hand motion tracking method for hand gesture interface using a camera in mobile devices such as a smart phone and PDA. When a camera moves according to the hand gesture of the user, global optical flows are generated. Therefore, robust hand movement estimation is possible by considering dominant optical flow based on histogram analysis of the motion direction. A continuous hand gesture is segmented into unit gestures by motion state estimation using motion phase, which is determined by velocity and acceleration of the estimated hand motion. Feature vectors are extracted during movement states and hand gestures are recognized at the end state of each gesture. Support vector machine (SVM), k-nearest neighborhood classifier, and normal Bayes classifier are used for classification. SVM shows 82% recognition rate for 14 hand gestures.