2019 Seventh International Symposium on Computing and Networking (CANDAR)最新文献

{"title":"GPGPU Implementation of Variational Bayesian Gaussian Mixture Models","authors":"Hiroki Nishimoto, Takashi Nakada, Y. Nakashima","doi":"10.1109/CANDAR.2019.00031","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00031","url":null,"abstract":"Machine learning technology has been developed along with the performance improvement of CPUs and GPGPUs, and it has been used for analyses of big data. Clustering is one task of them. Gaussian Mixture Models are a representation method that realizes clustering, and widely used in many applications for probability density modeling and soft clustering. However, the parameters of a GMM need to be estimated from data. There are some methods for estimate parameters, Variational Bayesian Gaussian Mixture Models (VB-GMM) are one of them, and it is hardly to over-fitting than other methods. But, it is computationally demanding. In this paper, we explain a parallel implementation of VB-GMM by GPGPU, and its evaluation. As a result, its execution speed is 524 times faster than that of CPU implementation.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"117 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132578945","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Extraction of Characteristic Subgraph Patterns with Support Threshold from Databases of Floor Plans","authors":"Tomonobu Ozaki","doi":"10.1109/CANDAR.2019.00033","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00033","url":null,"abstract":"The rental price of apartments is determined by various factors such as building age, floor space, and so forth. Although room layouts, or connection structures among rooms, are considered to affect the price, it is not obvious that what kind of layout has a strong influence. Identification of characteristic layouts on the price has a large impact on property lenders to determine appropriate rental price as well as on property borrowers to judge the reasonableness. In this paper, a framework to extract characteristic patterns on room layouts is proposed based on advanced techniques in frequent subgraph mining. Floor plans are represented as graphs, and class labels are given based on the estimated residuals by certain basic attributes. For the set of labeled floor plans, we develop a new class of contrast pattern having support threshold, under new definitions of support value of a pattern in a floor plan. In addition, we extract decision lists and decision sets consisting of significant subgraph patterns with thresholds by using methods in interpretability research of machine learning. The effectiveness of the proposed framework is evaluated in preliminary experiments using a real world dataset.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134079614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CANDAR 2019 Program Committee","authors":"Akihiro Fujiwara, Susumu Matsumae, Mattia D’emidio, Stéphane Devismes, M. Forsell, Masaru Fukushi, Genaro Juarez Martinez, Yonghwan Kim, Tomonobu Ozaki, P. Srimani, Daisuke Takafuji, Yasuhiko Takenaga, Jerry Trahan, Hiroshi Yamamoto, Jiangtao Yin, Eishi Arima, Naoki Fujieda, Kenji Kise, Kohta Nakashima, Y. Nakashima, Takeshi Ohkawa, K. Ootsu, Yasunori Osana, Shimpei Sato, Satoshi Fujita, Y. Kakuda, Masaki Aida, Takuya Asaka","doi":"10.1109/candar.2019.00007","DOIUrl":"https://doi.org/10.1109/candar.2019.00007","url":null,"abstract":"Track 1: Algorithms and Applications Chair: Akihiro Fujiwara, Kyushu Institute of Technology, Japan Chair: Susumu Matsumae, Saga University, Japan Mattia D’Emidio, University of L’Aquila, Italy Stéphane Devismes, VERIMAG UMR 5104, France Man Duhu, Shenzhen Institutes of Advanced Technology, China Martti Forsell, VTT, Finland Masaru Fukushi, Yamaguchi University, Japan Jing Gong, KTH Royal Institute of Technology, Sweden Teofilo Gonzalez, University of California, U.S. Qianping Gu, Simon Fraser University, Canada Teijiro Isokawa, University of Hyogo, Japan Chuzo Iwamoto, Hiroshima University, Japan Xiaohong Jiang, Future University Hakodate, Japan Genaro Juarez Martinez, University of the West of England, U.K. Hirotsugu Kakugawa, Ryukoku University, Japan Sayaka Kamei, Hiroshima University, Japan Yosuke Kikuchi, Tsuyama National College of Technology, Japan Yonghwan Kim, Nagoya Institute of Technology, Japan Anissa Lamani, École Internationale des Sciences du Traitement de l’Information, France Jia Lee, Chongqing University, China Yamin Li, Hosei University, Japan Luca Manzoni, Università degli Studi di Milano-Bicocca, Italy Ami Marowka, Bar-Ilan University, Israel Bruno Martin, Université Nice-Sophia Antipolis Gabriele Mencagli, University of Pisa, Italy Eiji Miyano, Kyushu Institute of Technology, Japan Hiroaki Mukaidani, Hiroshima University, Japan Takayuki Nagoya, Tottori University of Environmental Studies, Japan Kazuhiko Ohno, Mie University, Japan Hirotaka Ono, Nagoya University, Japan Fukuhito Ooshita, Nara Institute of Science and Technology, Japan Tomonobu Ozaki, Nihon University, Japan Max Plauth, Hasso Plattner Institute, Germany Jonas Posner, University of Kassel, Germany Victor Poupet, Université Montpellier 2, France Gokarna Sharma, Kent State University, U.S.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128388247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"2(1 – 1/ℓ)-Factor Steiner Tree Approximation in Õ(n^1/3) Rounds in the CONGESTED CLIQUE","authors":"Parikshit Saikia, S. Karmakar","doi":"10.1109/CANDAR.2019.00018","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00018","url":null,"abstract":"We study the Steiner tree problem in the CONGESTED CLIQUE model of distributed computing. We present a deterministic distributed approximation algorithm that computes a Steiner tree in Õ(n^1/3) rounds and Õ(n^7/3) messages for a given undirected weighted graph of n nodes with the approximation factor 2(1 - 1/ℓ) of the optimal, where ℓ is the number of terminal leaf nodes in the optimal Steiner tree. Note here that the Õ(⋅) notation hides polylogarithmic factors in n. To the best of our knowledge, this is the first work to study the Steiner tree problem in the CONGESTED CLIQUE model.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130487939","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An ANN Based Sequential Detection Method for Balancing Performance Indicators of IDS","authors":"Hao Zhao, Yaokai Feng, Hiroshi Koide, K. Sakurai","doi":"10.1109/CANDAR.2019.00039","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00039","url":null,"abstract":"In recent years, the number of cyber attacks has been increasing rapidly and network security has become an important issue. As a vital component of defense against network threats, intrusion detection system (IDS) was introduced and machine learning algorithms have been widely used in such systems for high detection performance. There are several evaluation indices such as false positive rate, false negative rate, and so on. A problem is that these indices are often related to each other. For example, while we try to decrease the false positive rate, the false negative rate often tends to increase, and vice versa. In this study, we proposed an ANN based sequential classifier method to mitigate this problem. Specifically, we try to train ANN to have a low false positive rate, despite which may lead to high false negative rate. Then, the reported negative instances are sent to the next ANN to make a further investigation, where the false negative instances reported at the previous ANN may be classified correctly. In this way, the final false negative rate can also be improved greatly. The results of the experiment shows that the proposed method can bring lower false negative rate and higher accuracy of detection while making the false positive rate at an acceptable level. Moreover, the optimum number of ANNs for our proposal is also investigated and discussed in this study.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122956251","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Copyright notice]","authors":"","doi":"10.1109/candar.2019.00003","DOIUrl":"https://doi.org/10.1109/candar.2019.00003","url":null,"abstract":"","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128964022","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"[Title page i]","authors":"","doi":"10.1109/candar.2019.00001","DOIUrl":"https://doi.org/10.1109/candar.2019.00001","url":null,"abstract":"","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128024245","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Folded Bloom Filter for High Bandwidth Memory, with GPU Implementations","authors":"Masato Hayashikawa, K. Nakano, Yasuaki Ito, Ryota Yasudo","doi":"10.1109/CANDAR.2019.00011","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00011","url":null,"abstract":"A Bloom filter is a space-efficient bit array data structure that can test if a query element is in a set (i.e. true) or not (i.e. false). It returns positive (i.e. the query element is in a set) or negative (i.e. the query element is not in a set). The Bloom filter may have false positive errors, because it returns positive for a false query. This paper focuses on the pattern test using Bloom filters such that a set P of strings (or patterns) is registered in bit arrays. For an input string T, the Bloom filter detects all substrings in T that are matched with one of the patterns in T. The main contribution of this paper is to present a new Bloom filter that we call folded Bloom filter and to implement it on the GPU for the pattern test. The folded Bloom filter is designed and implemented so that GPU memory access performance and parallelism are maximized. Our GPU implementation of the folded Bloom filter on NVIDIA Tesla V100 GPU can perform the pattern test for a set P with 744M (=744,261,120) patterns in 25.6Gbps throughput with false positive probability 1.20 × 10^-18. As far as we know, no previous Bloom filter implementation can attain such high throughput for a large set of patterns with quite small false positive probability.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128109518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Decomposing the Inverse of a Masked Vector in an Isomorphic Galois Field for Perfectly Masked S-Box","authors":"Yuta Kodera, Yuki Taketa, Takuya Kusaka, Y. Nogami","doi":"10.1109/CANDAR.2019.00027","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00027","url":null,"abstract":"The increment of opportunities for using machine learning (ML) technologies has brought a new threat to cryptosystems. As a remarkable example, the ML technologies have gradually been employed in the side-channel attack (SCA) to obtain sensitive information. In this paper, the authors focus on the structure of a masked S-Box in AES, which aims to equip the SCA resistance even for the attacks using the ML technologies. More precisely, this paper analyzes the mathematical structure of the inverse operation over F_(2^4)^2 which is an isomorphic field for obtaining efficient arithmetic for the AES, so that all functions in the encryption scheme can handle masked data as it is. The mathematical structure is realized by introducing several mathematical tools such as the Gauss periods and the Itoh-Tsujii inversion algorithm, and as a result, we clarified the factors of the coefficients of A^-1 for an element A ∊ F_(2^4)^2. It enables us to generate the corresponding element directly, which allows canceling the mask even after processing the SubBytes.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132130880","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Efficient GPU Implementations to Compute the Diameter of a Graph","authors":"Daisuke Takafuji, K. Nakano, Yasuaki Ito","doi":"10.1109/CANDAR.2019.00020","DOIUrl":"https://doi.org/10.1109/CANDAR.2019.00020","url":null,"abstract":"The Floyd-Warshall algorithm is a well-known algorithm to compute the distance of all pairs of nodes of a graph. The Blocked Floyd-Warshall algorithm, a variant of the Floyd-Warshall has been proposed to accelerate the Floyd-Warshall algorithm by means of a GPU architecture. The previously published GPU implementations for the Blocked Floyd-Warshall algorithm perform many separated kernel calls for costly barrier synchronization. The main contribution of this paper is to present a more sophisticated implementation of the Blocked Floyd-Warshall algorithm, which performs no barrier synchronization and invokes only one kernel call. Experimental results using NVIDIA Tesla V100 show that our implementation is 1.02-1.09 times faster than the previously published one. Our implementation with SIMD functions also runs 1.22-2.02 times faster than it. Furthermore, we propose efficient GPU implementations to execute the Blocked Floyd-Warshall algorithm for many graphs at the same time. From the experimental results, our implementation is 1.02-1.46 times faster than an extended implementation of previous one.","PeriodicalId":300306,"journal":{"name":"2019 Seventh International Symposium on Computing and Networking (CANDAR)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126953788","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}