Kyushu University Academic Staff Educational and Research Activities Database
List of Papers
Yoshinobu Kawahara Last modified date:2020.05.28

Professor / Division for Intelligent Societal Implementation of Mathmatical Computation / Institute of Mathematics for Industry


Papers
1. N. Uematsu, S. Umetani, and Y. Kawahara, An efficient branch-and-cut and heuristic algorithms for submodular function maximization, Journal of the Operations Research Society of Japan, 10.15807/jorsj.63.41, 63, 2, 41-59, 2020.05, [URL].
2. N. Takeishi, and Y. Kawahara, Knowledge-Based Regularization in Generative Modeling, Proc. of the 29th Int'l Joint Conf. on Artificial Intelligence and the 17th Pacific Rim Int'l Conf. on Artificial Intelligence (IJCAI-PRICAI'20), 2020.07.
3. H. Shiraishi, Y. Kawahara, R. Fukuma, O. Yamashita, S. Yamamoto, Y. Saitoh, H. Kishima, and T. Yanagisawa, Neural decoding of ECoG signals using dynamic mode decomposition, Journal of Neural Engineering, 10.1088/1741-2552/ab8910, 2020.04, [URL].
4. K. Fujii, N. Takeishi, M. Hojo, Y. Inaba, and Y. Kawahara, Physically-interpretable classification of network dynamics in complex collective motions, Scientific Reports, 10.1038/s41598-020-58064-w, 10, 3005, 2020.02, [URL].
5. I. Ul Haq, K. Fujii, and Y. Kawahara, Dynamic mode decomposition via dictionary learning for foreground modeling in videos, Proc. of the 15th Int'l Joint Conf. on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISAPP'20), 10.5220/0009144604760483, 476-483, 2020.03, [URL].
6. K. Fujii, N. Takeishi, B. Kibushi, M. Kouzaki, and Y. Kawahara, Data-driven spectral analysis for coordinative structures in periodic human locomotion, Scientific reports, 10.1038/s41598-019-53187-1, 9, 1, 2019.12, [URL].
7. N. Takeuchi, Y. Yoshida, and Y. Kawahara, Variational inference of penalized regression with submodular functions, Proc. of the 35th Conf. on Uncertainty in Artificial Intelligence (UAI'19), 443, 2019.10, [URL].
8. H. Yamashita, and Y. Kawahara, Principal points analysis via p-median problem for binary data, Journal of Applied Statistics, 10.1080/02664763.2019.1675605, 2019.01.
9. T. Bito, M. Hiraoka, and Y. Kawahara, Learning with Coherence Patterns in Multivariate Time-series Data via Dynamic Mode Decomposition, Proc. of the 2019 International Joint Conference on Neural Networks (IJCNN'19), 10.1109/IJCNN.2019.8852177, 2019.07, [URL].
10. K. Fujii, and Y. Kawahara, Dynamic mode decomposition in vector-valued reproducing kernel Hilbert spaces for extracting dynamical structure among observables, Neural Networks, 10.1016/j.neunet.2019.04.020, 117, 94-103, 2019.09, [URL].
11. K. Fujii, and Y. Kawahara, Supervised dynamic mode decomposition via multitask learning, Pattern Recognition Letters, 10.1016/j.patrec.2019.02.010, 122, 7-13, 2019.05, [URL].
12. M. Hojo, K. Fujii, and Y. Kawahara, Analysis of factors predicting who obtains a ball in basketball rebounding situations, International Journal of Performance Analysis in Sport, 10.1080/24748668.2019.1582892, 19, 2, 192-205, 2019.02, [URL].
13. M. Hojo, K. Fujii, Y. Inaba, Y. Motoyasu, and Y. Kawahara, Automatically recognizing strategic cooperative behaviors in various situations of a team sport, PLoS ONE, 10.1371/journal.pone.0209247, 13, 12, 2018.12, [URL].
14. I. Ishikawa, K. Fujii, M. Ikeda, Y. Hashimoto, and Y. Kawahara, Metric on nonlinear dynamical systems with Perron-Frobenius operators, Advances in Neural Information Processing Systems 31 (Proc. of NeurIPS'18), 2856-2866, 2018.12, [URL].
15. N. Takeishi, T. Yairo, and Y. Kawahara, Factorially-switching dynamic mode decomposition for Koopman analysis of time-variant systems, Proceedings of the 57th IEEE Conference on Decision and Control (CDC'18), 10.1109/CDC.2018.8619846, 6402-6408, 2018.12, [URL].
16. K. Fujii, T. Kawasaki, Y. Inaba, and Y. Kawahara, Prediction and classification in equation-free collective motion dynamics, PLoS Computational Biology, 10.1371/journal.pcbi.1006545, 14, 11, e1006545, 2018.11, [URL].
17. T. Wazawa, Y. Arai, Y. Kawahara, H. Takauchi, T. Washio, and T. Nagai, Highly biocompatible super-resolution fluorescence imaging using the fast photoswitching fluorescent protein Kohinoor and SPoD-ExPAN with Lp-regularized image reconstruction, Microscopy, 10.1093/jmicro/dfy004, 67, 2, 89-98, 2018.02, [URL].
18. N. Takeishi, Y. Kawahara, and T. Yairi, Learning Koopman invariant subspaces for dynamic mode decomposition, Advances in Neural Information Processing Systems 30 (Proc. of NIPS'17), 1131-1141, 2017.12, [URL].
19. K. Fujii, Y. Inaba, and Y. Kawahara, Koopman spectral kernels for comparing complex dynamics with application to multiagent in sports, Proceedings of the 2017 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD'17), 10.1007/978-3-319-71273-4_11, 127-139, 2017.12, [URL].
20. K. Takeuchi, Y. Kawahara, and T. Iwata, Structurally regularized non-negative tensor factorization for spatio-temporal pattern discoveries, Proceedings of the 2017 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD'17), 10.1007/978-3-319-71249-9_35, 582-598, 2017.12, [URL].
21. N. Takeishi, Y. Kawahara, and T. Yairi, Subspace dynamic mode decomposition for stochastic Koopman analysis, Physical Review E, 10.1103/PhysRevE.96.033310, 96, 033310, 2017.09, [URL].
22. N. Takeishi, Y. Kawahara, and T. Yairi, Sparse nonnegative dynamic mode decomposition, Proceedings of the 2017 IEEE International Conference on Image Processing (ICIP'17), 10.1109/ICIP.2017.8296769, 2682-2686, 2017.09, [URL].
23. N. Takeishi, Y. Kawahara, Y. Tabei, and T. Yairi, Bayesian dynamic mode decomposition, Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence (IJCAI'17), 10.24963/ijcai.2017/392, 2814-2821, 2017.07, [URL].
24. H. Wang, Y. Kawahara, C. Weng, and J. Yuan, Representative Selection with Structured Sparsity, Pattern Recognition, 10.1016/j.patcog.2016.10.014, 63, 268-278, 2017.03, [URL].
25. Y. Kawahara, Dynamic Mode Decomposition with Reproducing Kernels for Koopman Spectral Analysis, Advances in Neural Information Processing Systems 29 (Proc. of NIPS'16), 911-919, 2016.12, [URL].
26. B. Xin, Y. Kawahara, Y. Wang, L. Hu, and W. Gao, Efficient generalized fused lasso and its applications, ACM Transactions on Intelligent Systems and Technology, 10.1145/2847421, 7, 4, 2016.05, [URL].
27. M. Demeshko, T. Washio, Y. Kawahara, and Y. Pepyolyshev, A novel continuous and structural VAR modeling approach and its application to reactor noise analysis, ACM Transactions on Intelligent Systems and Technology, 10.1145/2710025, 7, 2, 2015.12, [URL].
28. Shinichi Yamagiwa, Yoshinobu Kawahara, Noriyuki Tabuchi, Yoshinobu Watanabe, Takeshi Naruo, Skill grouping method
Mining and clustering skill differences from body movement BigData, Proceedings of the 2015 IEEE International Conference on Big Data (IEEE Big Data 2015), 10.1109/BigData.2015.7364049, 2525-2534, 2015.12.
29. Takehide Hirata, Yoshinobu Kawahara, Masashi Sugiyama, Kazuya Asano, A fault detection technique for the steel manufacturing process based on a normal pattern library, Proceedings of the9th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFEPROCESS 2015), 10.1016/j.ifacol.2015.09.636, 28, 21, 871-876, 2015.09.
30. M. Demeshko, A. Dokhane, T. Washio, H. Ferroukhi, Yoshinobu Kawahara, C. Aguirre, Application of Continuous and Structural ARMA modeling for noise analysis of a BWR coupled core and plant instability event, Annals of Nuclear Energy, 10.1016/j.anucene.2014.08.045, 75, 645-657, 2015.01.
31. K. Takeuchi, Y. Kawahara, and T. Iwata, Higher order fused regularization for supervised learning with grouped parameters, Proc. of the European Conf. on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD'15), 10.1007/978-3-319-23528-8_36, 577-593, 2015.01, [URL].
32. Y. Kawahara, R. Iyer, and J.A. Bilmes, On approximate non-submodular minimization via tree-structured supermodularity, Proc. of the 18th Int'l Conf. on Artificial Intelligence and Statistics (AISTATS'15), 38, 444-452, 2015.01.
33. Keisuke Nagata, Takashi Washio, Yoshinobu Kawahara, Akira Unami, Toxicity prediction from toxicogenomic data based on class association rule mining, Toxicology Reports, 10.1016/j.toxrep.2014.10.014, 1, 1133-1142, 2014.12, While the recent advent of new technologies in biology such as DNA microarray and next-generation sequencer has given researchers a large volume of data representing genome-wide biological responses, it is not necessarily easy to derive knowledge that is accurate and understandable at the same time. In this study, we applied the Classification Based on Association (CBA) algorithm, one of the class association rule mining techniques, to the TG-GATEs database, where both toxicogenomic and toxicological data of more than 150 compounds in rat and human are stored. We compared the generated classifiers between CBA and linear discriminant analysis (LDA) and showed that CBA is superior to LDA in terms of both predictive performances (accuracy: 83% for CBA vs. 75% for LDA, sensitivity: 82% for CBA vs. 72% for LDA, specificity: 85% for CBA vs. 75% for LDA) and interpretability..
34. B. Xin, Y. Kawahara, Y. Wang, and W. Gao, Efficient generalized fused lasso and its application to the diagnosis of Alzheimer's disease, Proc. of the 28th AAAI Conf. on Artificial Intelligence (AAAI'14), 2163-2169, 2014.01.
35. M. Sugiyama, C.A. Azencott, D. Grimm, Y. Kawahara, and K.M. Borgwardt, Multi-task feature selection on multiple networks via maximum flows, Proc. of the 14th SIAM Int'l Conf. on Data Mining (SDM'14), 10.1137/1.9781611973440.23, 199-207, 2014.01, [URL].
36. Yunzhu Zheng, Haruka Suematsu, Takayuki Itoh, Ryohei Fujimaki, Satoshi Morinaga, Yoshinobu Kawahara, Scatterplot layout for high-dimensional data visualization, Journal of Visualization, 10.1007/s12650-014-0230-5, 18, 1, 111-119, 2014.01, Abstract: Multi-dimensional data visualization is an important research topic that has been receiving increasing attention. Several techniques that apply scatterplot matrices have been proposed to represent multi-dimensional data as a collection of two-dimensional data visualization spaces. Typically, when using the scatterplot-based approach it is easier to understand relations between particular pairs of dimensions, but it often requires too large display spaces to display all possible scatterplots. This paper presents a technique to display meaningful sets of scatterplots generated from high-dimensional datasets. Our technique first evaluates all possible scatterplots generated from high-dimensional datasets, and selects meaningful sets. It then calculates the similarity between arbitrary pairs of the selected scatterplots, and places relevant scatterplots closer together in the display space while they never overlap each other. This design policy makes users easier to visually compare relevant sets of scatterplots. This paper presents algorithms to place the scatterplots by the combination of ideal position calculation and rectangle packing algorithms, and two examples demonstrating the effectiveness of the presented technique..
37. Haruka Suematsu, Zheng Yunzhu, Takayuki Itoh, Ryohei Fujimaki, Satoshi Morinaga, Yoshinobu Kawahara, Arrangement of low-dimensional parallel coordinate plots for high-dimensional data visualization, 2013 17th International Conference on Information Visualisation, IV 2013 Proceedings - 2013 17th International Conference on Information Visualisation, IV 2013, 10.1109/IV.2013.7, 59-65, 2013.12, Multidimensional data visualization is an important research topic that has been receiving increasing attention. Several techniques that use parallel coordinate plots have been proposed to represent all dimensions of data in a single display space. In addition, several other techniques that apply scatter plot matrices have been proposed to represent multidimensional data as a collection of low-dimensional data visualization spaces. Typically, when using the latter approach it is easier to understand relations among particular dimensions, but it is often difficult to observe relations between dimensions separated into different visualization spaces. This paper presents a framework for displaying an arrangement of low-dimensional data visualization spaces that are generated from high-dimensional datasets. Our proposed technique first divides the dimensions of the input datasets into groups of lower dimensions based on their correlations or other relationships. If the groups of lower dimensions can be visualized in independent rectangular spaces, our technique packs the set of low-dimensional data visualizations into a single display space. Because our technique places relevant low-dimensions closer together in the display space, it is easier to visually compare relevant sets of low-dimensional data visualizations. In this paper, we describe in detail how we implement our framework using parallel coordinate plots, and present several results demonstrating its effectiveness..
38. Y. Sogawa, T. Ueno, Y. Kawahara, and T. Washio, Active learning for noisy oracle via density power divergence, Neural Networks, 10.1016/j.neunet.2013.05.007, 46, 133-143, 2013.10, [URL].
39. C.A. Azencott, D. Grimm, M. Sugiyama, Y. Kawahara, and K.M. Borgwardt, Efficient network-guided multi-locus association mapping with graph cuts, Bioinformatics, 10.1093/bioinformatics/btt238, 29, 13, 2013.07, [URL].
40. K. Nagano, and Y. Kawahara, Structured convex optimization under submodular constraints, Proc. of the 29th Ann. Conf. on Uncertainty in Artificial Intelligence (UAI'13), 459-468, 2013.07, [URL].
41. Yasuhiro Sogawa, Tsuyoshi Ueno, Yoshinobu Kawahara, Takashi Washio, Active learning for regression via density power divergence, Transactions of the Japanese Society for Artificial Intelligence, 10.1527/tjsai.28.13, 28, 1, 13-21, 2013.01, The accuracy of active learning is crucially influenced by the existence of noisy labels given by a real-world noisy oracle. In this paper, we propose a novel pool-based active learning framework through density power divergence. It is known that density power divergence, such as β-divergence and γ-divergence, can be accurately estimated even under the existence of outliers (noisy labels) within data. In addition, we propose an evaluation scheme for these measures based on those asymptotic statistical analyses, which enables us to perform active learning by evaluating an estimation variance. Experiments on artificial and real-world datasets show that our active learning scheme performs better than state-of-the-art methods..
42. A. Takeda, M. Niranjan, J. Gotoh, and Y. Kawahara, Simultaneous pursuit of out-of-sample performance and sparsity in index tracking portfolios, Computational Management Science, 10.1007/s10287-012-0158-y, 10, 1, 21-49, 2013.01, [URL].
43. T. Ueno, K. Hayashi, T. Washio, and Y. Kawahara, Weighted likelihood policy search with model selection, Advances in Neural Information Processing Systems 25 (Proc. of NIPS'12), 2357-2365, 2012.12.
44. Yasuhiro Sogawa, Tsuyoshi Ueno, Yoshinobu Kawahara, Takashi Washio, Robust active learning for linear regression via density power divergence, 19th International Conference on Neural Information Processing, ICONIP 2012 Neural Information Processing - 19th International Conference, ICONIP 2012, Proceedings, 10.1007/978-3-642-34487-9_72, 594-602, 2012.11, The performance of active learning (AL) is crucially influenced by the existence of outliers in input samples. In this paper, we propose a robust pool-based AL measure based on the density power divergence. It is known that the density power divergence can be accurately estimated even under the existence of outliers within data. We further derive an AL scheme based on an asymptotic statistical analysis on the M-estimator. The performance of the proposed framework is investigated empirically using artificial and real-world data..
45. S. Hara, Y. Kawahara, T. Washio, P. von Bünau, T. Tokunaga, and K. Yumoto, Separation of stationary and non-stationary sources with a generalized eigenvalue problem, Neural Networks, 10.1016/j.neunet.2012.04.001, 33, 7-20, 2012.09, [URL].
46. Y. Kawahara, and M. Sugiyama, Sequential change-point detection based on direct density-ratio estimation, Statistical Analysis and Data Mining, 10.1002/sam.10124, 5, 2, 114-127, 2012.04, [URL].
47. Y. Kawahara, and T. Washio, Prismatic algorithm for discrete D.C. programming problem, Advances in Neural Information Processing Systems 24 (Proc. of NIPS'11), 2011.12.
48. Masao Joko, Yoshinobu Kawahara, Takehisa Yairi, Learning non-linear dynamical systems by alignment of local linear models, Transactions of the Japanese Society for Artificial Intelligence, 10.1527/tjsai.26.638, 26, 6, 638-648, 2011.10, In this paper, we present an algorithm for learning non-linear dynamical systems which works by aligning local linear models, based on a probabilistic formulation of subspace identification. This is achieved by the fusion of the recent works in the fields of machine learning and system control. Because the procedure for constructing a state sequence in subspace identification can be interpreted as the Canonical Correlation Analysis(CCA) between past and future observation sequences, we can derive a latent variable representation for this problem. Therefore, as in a similar manner to the recent works on learning a mixture of probabilistic models, we obtain a framework for constructing a state space by aligning local linear coordinates. This leads to a prominent algorithm for learning non-linear dynamical systems. Finally, we apply our method to motion capture data and telemetry data, and then show how our algorithm works well..
49. K. Nagano, Y. Kawahara, and K. Aihara, Size-constrained submodular minimization through minimum norm base, Proc. of the 28th International Conference on Machine Learning (ICML'11), 977-984, 2011.10.
50. Masao Joko, Yoshinobu Kawahara, Takehisa Yairi, Learning non-linear dynamical systems by alignment of local linear models, Transactions of the Japanese Society for Artificial Intelligence, 10.1527/tjsai.26.638, 26, 6, 638-648, 2011.10, In this paper, we present an algorithm for learning non-linear dynamical systems which works by aligning local linear models, based on a probabilistic formulation of subspace identification. This is achieved by the fusion of the recent works in the fields of machine learning and system control. Because the procedure for constructing a state sequence in subspace identification can be interpreted as the Canonical Correlation Analysis(CCA) between past and future observation sequences, we can derive a latent variable representation for this problem. Therefore, as in a similar manner to the recent works on learning a mixture of probabilistic models, we obtain a framework for constructing a state space by aligning local linear coordinates. This leads to a prominent algorithm for learning non-linear dynamical systems. Finally, we apply our method to motion capture data and telemetry data, and then show how our algorithm works well..
51. Y. Kawahara, S. Shimizu, and T. Washio, Analyzing relationships among ARMA processes based on non-Gaussianity of external influences, Neurocomputing, 10.1016/j.neucom.2011.02.008, 74, 12-13, 2212-2221, 2011.06, [URL].
52. T. Inazumi, T. Washio, S. Shimizu, J. Suzuki, A. Yamamoto, and Y. Kawahara, Discovering causal structures in binary exclusive-or skew acyclic models, Proc. of the 27th Ann. Conf. on Uncertainty in Artificial Intelligence (UAI'11), 373-382, 2011.07.
53. S. Shimizu, T. Inazumi, Y. Sogawa, A. Hyvärinen, Y. Kawahara, T. Washio, P.O. Hoyer, and K. Bollen, DirectLiNGAM: A direct method for learning a linear non-gaussian structural equation model, Journal of Machine Learning Research, 12, 1225-1248, 2011.04.
54. Y. Kawahara, K. Nagano, and Y. Okamoto, Submodular fractional programming for balanced clustering, Pattern Recognition Letters, 10.1016/j.patrec.2010.08.008, 32, 2, 235-243, 2011.01, [URL].
55. Yasuhiro Sogawa, Shohei Shimizu, Yoshinobu Kawahara, Takashi Washio, An experimental comparison of linear non-Gaussian causal discovery methods and their variants, 2010 6th IEEE World Congress on Computational Intelligence, WCCI 2010 - 2010 International Joint Conference on Neural Networks, IJCNN 2010 2010 IEEE World Congress on Computational Intelligence, WCCI 2010 - 2010 International Joint Conference on Neural Networks, IJCNN 2010, 10.1109/IJCNN.2010.5596737, 2010.12, Many multivariate Gaussianity-based techniques for identifying causal networks of observed variables have been proposed. These methods have several problems such that they cannot uniquely identify the causal networks without any prior knowledge. To alleviate this problem, a non-Gaussianity-based identification method LiNGAM was proposed. Though the LiNGAM potentially identifies a unique causal network without using any prior knowledge, it needs to properly examine independence assumptions of the causal network and search the correct causal network by using finite observed data points only. On another front, a kernel based independence measure that evaluates the independence more strictly was recently proposed. In addition, some advanced generic search algorithms including beam search have been extensively studied in the past. In this paper, we propose some variants of the LiNGAM method which introduce the kernel based method and the beam search enabling more accurate causal network identification. Furthermore, we experimentally characterize the LiNGAM and its variants in terms of accuracy and robustness of their identification..
56. K. Nagano, Y. Kawahara, and S. Iwata, Minimum average cost clustering, Advances in Neural Information Processing Systems 23 (NIPS'10), 2010.12.
57. Satoshi Hara, Yoshinobu Kawahara, Takashi Washio, Paul Von Bünau, Stationary subspace analysis as a generalized eigenvalue problem, 17th International Conference on Neural Information Processing, ICONIP 2010 Neural Information Processing Theory and Algorithms - 17th International Conference, ICONIP 2010, Proceedings, 10.1007/978-3-642-17537-4_52, 422-429, 2010.12, Understanding non-stationary effects is one of the key challenges in data analysis. However, in many settings the observation is a mixture of stationary and non-stationary sources. The aim of Stationary Subspace Analysis (SSA) is to factorize multivariate data into its stationary and non-stationary components. In this paper, we propose a novel SSA algorithm (ASSA) that extracts stationary sources from multiple time series blocks. It has a globally optimal solution under certain assumptions that can be obtained by solving a generalized eigenvalue problem. Apart from the numerical advantages, we also show that compared to the existing method, fewer blocks are required in ASSA to guarantee the identifiability of the solution. We demonstrate the validity of our approach in simulations and in an application to domain adaptation..
58. Takehisa Yairi, Minoru Inui, Akihiro Yoshiki, Yoshinobu Kawahara, Noboru Takata, Spacecraft telemetry data monitoring by dimensionality reduction techniques, Proceedings of SICE Annual Conference 2010, SICE 2010 - Final Program and Papers, 1230-1234, 2010.01, In this paper, we consider a "data-driven" anomaly detection framework for spacecraft systems using dimensionality reduction and reconstruction techniques. This method first learns a mapping from the original data space to a low dimensional space and its reverse mapping by applying linear or non-linear dimensionality reduction algorithms to a normal training data set. After the training, it applies the learned pair of mappings to a test data set to obtain a reconstructed data set, and then evaluate the reconstruction errors. We will show the results of applying several representative linear and non-lineardimensionality reduction algorithms with this framework to the electrical power subsystem (EPS) data of actual artificial satellites..
59. Masao Joko, Yoshinobu Kawahara, Takehisa Yairi, Learning non-linear dynamical systems by alignment of local linear models, 2010 20th International Conference on Pattern Recognition, ICPR 2010 Proceedings - 2010 20th International Conference on Pattern Recognition, ICPR 2010, 10.1109/ICPR.2010.271, 1084-1087, 2010.11, Learning dynamical systems is one of the important problems in many fields. In this paper, we present an algorithm for learning non-linear dynamical systems which works by aligning local linear models, based on a probabilistic formulation of subspace identification. Because the procedure for constructing a state sequence in subspace identification can be interpreted as the CCA between past and future observation sequences, we can derive a latent variable representation for this problem. Therefore, as in a similar manner to the recent works on learning a mixture of probabilistic models, we obtain a framework for constructing a state space by aligning local linear coordinates. This leads to a prominent algorithm for learning non-linear dynamical systems. Finally, we apply our method to motion capture data and show how our algorithm works well..
60. Y. Kawahara, and M. Sugiyama, Change-point detection in time-series data by direct density-ratio estimation, Proc. of the 9th SIAM Int'l Conf. on Data Mining (SDM'09), 385-396, 2009.12.
61. Y. Kawahara, K. Nagano, K. Tsuda, and J.A. Bilmes, Submodularity cuts and applications, Advances in Neural Information Processing Systems 22 (Proc. of NIPS'09), 916-924, 2009.12, [URL].
62. S. Shimizu, A. Hyvarinen, Y. Kawahara, and T. Washio, A direct method for estimating a causal ordering in a linear non-Gaussian acyclic model, Proc. of the 25th Ann. Conf. on Uncertainty in Artificial Intelligence (UAI'09), 506-513, 2009.07.
63. Yoshinobu Kawahara, Takehisa Yairi, Kazuo Machida, Change-point detection algorithms based on subspace methods, Transactions of the Japanese Society for Artificial Intelligence, 23, 2, 76-84, 2008.04, In this paper, we propose a class of algorithms for detecting the change-points in time-series data based on subspace identification, which is originaly a geometric approach for estimating linear state-space models generating time-series data. Our algorithms are derived from the principle that the subspace spanned by the columns of an observability matrix and the one spanned by the subsequences of time-series data are approximately equivalent. In this paper, we derive a batch-type algorithm applicable to ordinary time-series data, i.e., consisting of only output series, and then introduce the online version of the algorithm and the extension to be available with input-output time-series data. We illustrate the superior performance of our algorithms with comparative experiments using artificial and real datasets..
64. Y. Kawahara, T. Yairi, and K. Machida, Change-point detection in time-series data based on subspace identification, Proceedings of the 7th IEEE International Conference on Data Mining (ICDM'07), 10.1109/ICDM.2007.78, 559-564, 2007.12, [URL].
65. Y. Kawahara, T. Yairi, and K. Machida, A kernel subspace method by stochastic realization for learning nonlinear dynamical systems, Advances in Neural Information Processing Systems 19 (Proc. of NIPS'06), 665-672, 2007.12, [URL].
66. Rei Fujiki, Hideyuki Tanaka, Yoshinobu Kawahara, Takehisa Yairi, Kazuo Machida, Autonomous recognition of multiple cable topology with image, 2006 SICE-ICASE International Joint Conference 2006 SICE-ICASE International Joint Conference, 10.1109/SICE.2006.315754, 1425-1430, 2006.12, From industrial situation to ordinary life, there are a lot of cables in today's society. There, however, are not many studies that describe the topological relationship between multibple cables for autonomous operation of robotics device. In this paper, an autonomous computing way to create 3-D topological model of multiple cables using image is discussed. Cable model is described by probablity density function and optimized its parameters via EM Algorithm, that indicates the configuration and topological relationships. In order to manipulate the object cables then, depth from sensor to points are added to the unique points in topological graph structure derived from EM model. The quantitative results of basic experiment that was aimed to prove the recognition possibility via the algorithm are reported..
67. Takehisa Yairi, Yoshinobu Kawahara, Ryohei Fujimaki, Yuichi Sato, Kazuo Machida, Telemetry-mining
A machine learning approach to anomaly detection and fault diagnosis for space systems, SMC-IT 2006: 2nd IEEE International Conference on Space Mission Challenges for Information Technology Proceedings - SMC-IT 2006 2nd IEEE International Conference on Space Mission Challenges for Information Technology, 10.1109/SMC-IT.2006.79, 466-473, 2006.12, For any space mission, safety and reliability are the most important issues. To tackle this problem, we have studied anomaly detection and fault diagnosis methods for spacecraft systems based on machine learning (ML) and data mining (DM) technology. In these methods, the knowledge or model which is necessary for monitoring a spacecraft system is (semi-)automatically acquired from the spacecraft telemetry data. In this paper, we first overview the anomaly detection / diagnosis problem in the spacecraft systems and conventional techniques such as limit-check, expert systems and model-based diagnosis. Then we explain the concept of ML/DM-based approach to this problem, and introduce several anomaly detection / diagnosis methods which have been developed by us..
68. Yuichi Sato, Yoshinobu Kawahara, Takehisa Yairi, Kazuo Machida, Visualization of spacecraft data based on interdependency between changing points in time series, 2006 SICE-ICASE International Joint Conference 2006 SICE-ICASE International Joint Conference, 10.1109/SICE.2006.315124, 3414-3418, 2006.12, A support technology for spacecraft operators is one of the important themes for reliable operation. We suggest a framework for visualization of relations among sequences based on "changing points". First, we employ auto-regression model for detecting changing points from data. And next, we apply a structure learning of dynamic Bayesian Net to the change-detected data for getting the graph structure, which stands for dependency among sequences. We applied this approach to two kinds of actual telemetry data of a communication satellite, and verified graph structures rightly showed the relation among sequences..
69. Spacecraft diagnosis method using dynamic Bayesian networks.
70. Yoshinobu Kawahara, Takehisa Yairi, Kazuo Machida, Diagnosis method for spacecraft using dynamic bayesian networks, i- SAIRAS 2005 - The 8th International Symposium on Artificial Intelligence, Robotics and Automation in Space European Space Agency, (Special Publication) ESA SP, 603, 649-656, 2005.12, Development of sophisticated anomaly detection and diagnosis methods for spacecraft is one of the important problems in space system operation. In this study, we propose a diagnosis method for spacecraft using probabilistic reasoning and statistical learning with Dynamic Bayesian Networks (DBNs). In this method, the DBNs are initially from prior-knowledge, then modified or partly re-constructed by statistical learning with operation data, as a result adaptable and in-depth diagnosis is performed by probabilistic reasoning using the DBNs. The proposed method was applied to the telemetry data that simulates the malfunction of thrusters in rendezvous maneuver of spacecraft, and the effectiveness of the method was confirmed..
71. Rei Fujiki, Hideyuki Tanaka, Yoshinobu Kawahara, Takehisa Yairi, Kazuo Machida, Autonomous recognition of multiple cable topology with image, 2006 SICE-ICASE International Joint Conference 2006 SICE-ICASE International Joint Conference, 10.1109/SICE.2006.315754, 1425-1430, 2006.12, From industrial situation to ordinary life, there are a lot of cables in today's society. There, however, are not many studies that describe the topological relationship between multibple cables for autonomous operation of robotics device. In this paper, an autonomous computing way to create 3-D topological model of multiple cables using image is discussed. Cable model is described by probablity density function and optimized its parameters via EM Algorithm, that indicates the configuration and topological relationships. In order to manipulate the object cables then, depth from sensor to points are added to the unique points in topological graph structure derived from EM model. The quantitative results of basic experiment that was aimed to prove the recognition possibility via the algorithm are reported..
72. Takehisa Yairi, Yoshinobu Kawahara, Ryohei Fujimaki, Yuichi Sato, Kazuo Machida, Telemetry-mining
A machine learning approach to anomaly detection and fault diagnosis for space systems, SMC-IT 2006: 2nd IEEE International Conference on Space Mission Challenges for Information Technology Proceedings - SMC-IT 2006 2nd IEEE International Conference on Space Mission Challenges for Information Technology, 10.1109/SMC-IT.2006.79, 466-473, 2006.12, For any space mission, safety and reliability are the most important issues. To tackle this problem, we have studied anomaly detection and fault diagnosis methods for spacecraft systems based on machine learning (ML) and data mining (DM) technology. In these methods, the knowledge or model which is necessary for monitoring a spacecraft system is (semi-)automatically acquired from the spacecraft telemetry data. In this paper, we first overview the anomaly detection / diagnosis problem in the spacecraft systems and conventional techniques such as limit-check, expert systems and model-based diagnosis. Then we explain the concept of ML/DM-based approach to this problem, and introduce several anomaly detection / diagnosis methods which have been developed by us..
73. Yuichi Sato, Yoshinobu Kawahara, Takehisa Yairi, Kazuo Machida, Visualization of spacecraft data based on interdependency between changing points in time series, 2006 SICE-ICASE International Joint Conference 2006 SICE-ICASE International Joint Conference, 10.1109/SICE.2006.315124, 3414-3418, 2006.12, A support technology for spacecraft operators is one of the important themes for reliable operation. We suggest a framework for visualization of relations among sequences based on "changing points". First, we employ auto-regression model for detecting changing points from data. And next, we apply a structure learning of dynamic Bayesian Net to the change-detected data for getting the graph structure, which stands for dependency among sequences. We applied this approach to two kinds of actual telemetry data of a communication satellite, and verified graph structures rightly showed the relation among sequences..