Search Results 1-20 of 1003

  • Fast Nearest Neighbor Search with Narrow 16-bit Sketch  [in Japanese]

    樋口 直哉 , 今村 安伸 , 久保山 哲二 , 平田 耕一 , 篠原 武

    … ためには,32ビットスケッチを用いる場合より大きな候補数Kを必要とする.データオブジェクトをスケッチの値によってソートしておくことで,第2段階の検索におけるメモリ局所性を向上することで候補数Kの増加による速度低下を低減できる.提案手法を用いると,検索精度を維持しつつ10倍程度の高速化が実現できる.We discuss k nearest neighbor search using sketches, which is a kind of locality sensitive hash …

    情報処理学会論文誌数理モデル化と応用(TOM) 13(1), 13-22, 2020-03-25

    IPSJ 

  • Success/Failure Prediction for Final Examination Using the LCT Results and FPT Results  [in Japanese]

    廣瀬 英雄

    … The author proposed a prediction method to find students at drop-out risks as early as possible byusing the learning check testing (LCT) results, where the nearest-neighbor method was used to findthe similarity of the cumulative LCT ability trends in the previous paper. …

    広島工業大学紀要. 研究編 54, 133-136, 2020-02

    IR 

  • Success/Failure Prediction for Final Examination Using the LCT Results : A Model using 2017 LCT Data is Applied to 2018 LCT Data  [in Japanese]

    廣瀬 英雄

    … The author proposed a prediction method to find students with drop-out risks as early as possibleby using the learning check testing (LCT) results, where the nearest-neighbor method was used inthe previous paper. … Thus, in this paper,we investigated whether the proposed method was valid or not by applying the model built using2017 LCT data to newly obtained 2018 LCT results. …

    広島工業大学紀要. 研究編 54, 137-140, 2020-02

    IR 

  • A SLOPE FAILURE TYPE CLASSIFICATION ALGORITHM FOR MULTIPLE SIMULTANEOUS SLOPE FAILURE HAZARD ASSESSMENT  [in Japanese]

    UCHIDA Ikko , KOJIMA Hirohito

    … <p> Pointing out the necessity of evaluating hazardous areas affected by different types of multiple simultaneous slope failures, this paper proposes a slope failure type classification algorithm based on multigroup discriminant analysis of the Hayashi's quantification method type II. … The scatter diagram based on multigroup quantification method type II indicates that the training samples can be grouped into three classes corresponding to above slope failure types. …

    Journal of Japan Society of Civil Engineers, Ser. F3 (Civil Engineering Informatics) 76(1), 42-52, 2020

    J-STAGE 

  • An Anomaly Detection Method with Neural Network Near Neighbor  [in Japanese]

    KATO Yuichi , TAKAGI Kentaro , TAO Yaling , NAITO Susumu , TAGUCHI Yasunori , BUDIANTO Teguh , NAKATA Kouta

    … k-nearest neighbor and AutoEncoder are one of those methods used frequently. … However, the former one searches nearest neighbors even for anomalies requiring large distance between normal and anomaly examples, and the latter one suffers from excessive generalization. … In order to tackle the problems, we developed an algorithm, neural network near neighbor, approximating neighborhood search by a neural network. …

    Proceedings of the Annual Conference of JSAI JSAI2020(0), 2I4GS202-2I4GS202, 2020

    J-STAGE 

  • Exploring the Characteristics in Spatial Patterns of Haunted Spots using Nearest-neighbor Spatial-association Measure  [in Japanese]

    Suzuki Koshiro , Ito Shuichi , U Yennan

    … <p>Nearest-neighbor Spatial-association Measure (NSM), a method for analyzing the spatial relationship between two different types of point distributions, has applied to investigate the contingency of two different point distribution tendencies such as chain stores, firms and convenient stores. …

    Proceedings of the General Meeting of the Association of Japanese Geographers 2020s(0), 31, 2020

    J-STAGE 

  • A Kernel <i>k</i>-Means-Based Method and Attribute Selections for Diabetes Diagnosis

    Cao Tru , Vo Chau , Nguyen Son , Inoue Atsushi , Zhou Duanning

    … First, we propose a kernel <i>k</i>-means-based prediction method and explore attribute selections for effective and robust diabetes diagnosis. … This method derives homogeneous sub-clusters in the high dimensional kernelized feature space to compute the distance of a new instance to those sub-clusters, and then apply the 1-nearest neighbor to classify it as positive or negative to the disease. …

    Journal of Advanced Computational Intelligence and Intelligent Informatics 24(1), 73-82, 2020

    J-STAGE 

  • Full-Potential KKR Calculations for Interaction Energies in Al-Rich AlX (X = H∼Sn) Alloys: I. Fundamental Features and Thermal Electronic Contribution due to Fermi-Dirac Distribution

    Asato Mitsuhiro , Liu Chang , Fujima Nobuhisa , Hoshino Toshiharu , Mohri Tetsuo

    … = 1∼4) interaction energies (IEs) in Al-rich AlX (X = H∼Sn) alloys, by using the full-potential Korringa-Kohn-Rostoker Green's function (FPKKR) method, and clarified the fundamental features and the thermal electronic contribution due to the Fermi Dirac (FD) distribution for these IEs. … impurities are strongly repulsive at the 1st-nearest neighbor (<i>nn</i>) and show the oscillating behavior with the interatomic distance; …

    MATERIALS TRANSACTIONS 61(1), 94-103, 2020

    J-STAGE 

  • Classification Based on Skyline Objects : Preliminary Report  [in Japanese]

    大久保 好章

    電気学会研究会資料. IS 2019(46-51・53-58), 1-6, 2019-08-29

  • A Method of Describing a Self-occlusive Motion - A Reverse Motion History Image

    Tan Joo Kooi , Okae Sayaka , Yamashita Youtaro , Ono Yuta

    … This paper proposes a new method of describing a self-occlusive human motion, particularly in the depth direction, which has been considered little in the motion/action recognition studies to date in spite of its importance in our daily life. … A Motion History Image (MHI) is a well-known method of describing a motion by a single gray value image, but it suffers from a self-occlusion problem in which present motion overwrites past motion. …

    International Journal of Biomedical Soft Computing and Human Sciences 24(1), 1-7, 2019-07

    IR 

  • Ensemble Feature Learning to Identify Risk Factors for Predicting Secondary Cancer

    叶 秀彩 , 櫻井 鉄也 , Xiucai YE , Hongmin Li , Tetsuya SAKURAI , Pei-Wei Shueng

    … Methods: We propose an effective ensemble feature learning method to identify the risk factors for predicting secondary cancer by considering class imbalance and patient heterogeneity. … In each group, we apply the oversampling method to balance the number of samples in each class and use them as training data for ensemble feature learning. …

    International journal of medical sciences 16(7), 949-959, 2019-06

    IR 

  • Fast and Diversity-aware Spatio-temporal Search for Twitter using Pre-search  [in Japanese]

    落合 桂一 , 鳥居 大祐 , 深澤 佑介 , 松尾 豊

    … K-Nearest Neighbor query can search quickly, however, there is diversity problem. … Therefore, we propose a fast and diversity-aware method for searching nearby tweets using location adaptive range query. … Our evaluation on one million tweets shows that the proposed method is about 10.1%-29.0% faster than the diversity aware baselines. …

    情報処理学会論文誌 60(2), 549-560, 2019-02-15

    IPSJ 

  • Emotion Recognition for Japanese Short Sentences Including Slangs Based on Bag of Concepts Feature Trained by Large Web Text

    Matsumoto Kazuyuki , Yoshida Minoru , Kita Kenji

    … We prepared an emotion corpus by collecting weblog article texts including new words, analyzed the corpus statistically, and proposed a method to estimate emotions of the texts. … We defined this method as a method using Bag of Concepts as feature. … As a result of the evaluation experiment using several classifiers, the proposed method was proved its robustness for unknown expressions. …

    Current Analysis on Instrumentation and Control 2019(2), 9-18, 2019-02-01

    IR 

  • A DEEP-LEARNING METHOD TO CORRECT THE OUTPUT OF A NUMERICAL WEATHER PREDICTION MODEL USED FOR SHORT-TERM PRECIPITATION FORECAST  [in Japanese]

    KURAKAMI Takeru , SOUMA Kazuyoshi , MIYAMOTO Takashi , FURUYA Takahiko , MAGOME Jun , ISHIDAIRA Hiroshi

    … To correct the numerical weather prediction model output, it might be effective to apply deep learning techniques, which have led to remarkable achievements in processing big data.</p><p> In this study, we developed a deep learning method using a convolution neural network (CNN) to correct numerical weather prediction model outputs. …

    Journal of Japan Society of Civil Engineers, Ser. G (Environmental Research) 75(5), I_33-I_39, 2019

    J-STAGE 

  • Atomic and Effective Pair Interactions in FeC Alloy with Point Defects: A Cluster Expansion Study

    Nguyen Tien Quang , Escano Mary Clare Sison , Sato Kazunori , Shibutani Yoji , Oguchi Tamio , Mohri Tetsuo

    … <p>The Cluster Expansion Method (CEM) is used to investigate the pair interactions in body centered cubic (BCC) FeC alloy in the presence of vacancies. … We found that, in this alloy, the interaction mostly comes from the first nearest neighbor pairs, and, to some degree, from the third nearest neighbor pairs. …

    ISIJ International 59(12), 2343-2351, 2019

    J-STAGE 

  • Energy Minimization over <i>m</i>-Branched Enumeration for Generalized Linear Subspace Clustering

    ZHANG Chao

    … To settle this problem, we take advantages of both greedy strategy and energy minimization strategy to propose a simple yet effective algorithm based on the assumption that an <i>m</i>-branched (i.e., perfect <i>m</i>-ary) tree which is constructed by collecting <i>m</i>-nearest neighbor points in each node has a high probability of containing the near-exact subspace. …

    IEICE Transactions on Information and Systems E102.D(12), 2485-2492, 2019

    J-STAGE 

  • <b>Fault Detection for Railway Traction Motor Bearing through Leakage Current</b>

    SAKAIDANI Yo , KONDO Minoru , TAKAHASHI Ken

    … In this paper, a bearing fault detection method for a railway traction motor through leakage currents is proposed. … The proposed detection method combines octave band analysis and machine learning. … Experiments simulating abnormalities due to defective bearings were conducted and the effectiveness of the proposed method was verified. …

    Quarterly Report of RTRI 60(4), 256-261, 2019

    J-STAGE 

  • Estimation Accuracy of Partial <i>Q-Q</i> Plot of Log-Weibull Distribution  [in Japanese]

    Sayama Shuji

    … Fit the approximate line to this point and the nearest ones on both sides by least squares method. … Adding a neighbor point on the right side of these ones, fit it similarly, and then adding a neighbor point on the left side of these ones, fit it similarly. … plot is more rigorous fit of the distribution to the data than maximum log-likelihood method.</p> …

    IEEJ Transactions on Fundamentals and Materials 139(11), 596-602, 2019

    J-STAGE 

  • Impact Evaluation Model of the Effects of Planned Farmland Accumulation on Traveling Efficiency in Paddy Field Farms  [in Japanese]

    Morisawa Kensaku , Yamashita Ryohei

    … The optimal route was identified by a combination of the nearest neighbor method and the 2-opt method. …

    Agricultural Information Research 28(3), 115-126, 2019

    J-STAGE 

  • Scalable Community Identification with Manifold Learning on Speaker I-Vector Space

    WANG Hongcui , LIU Shanshan , JIN Di , LI Lantian , DANG Jianwu

    … For this problem, we consider to convert the speaker i-vector representation of utterances in the Euclidean space into a network structure constructed based on the local (<i>k</i>) nearest neighbor relationship of these signals. … The results provided the insight into the structure of the speakers present in the data and also confirmed the effectiveness of the proposed new method. …

    IEICE Transactions on Information and Systems E102.D(10), 2004-2012, 2019

    J-STAGE 

Page Top