Research interests : Statistical Learning Theory, Machine Learning, Computer Vision, Data Mining
Email : nakajima(at) --- Please replace (at) with "@".


- Our papers have been accepted:

A. Bauer, S. Nakajima, N. Goernitz, K.-R. Muller, ''Partial Optimality of Dual Decomposition for MAP Inference in Pairwise MRFs,''
AISTATS 2019, (Naha, Japan, April 16-18, 2019).

S. Kaltenstadler, S. Nakajima, K-R. Mueller, W. Samek, ''Wasserstein Stationary Subspace Analysis,''
IEEE Journal of Selected Topics in Signal Processing, vol.12(6), pp.1213-1223, 2018.

W. Pronobis, D. Panknin, J. Kirschnick, V. Srinivasan, W. Samek, V. Markl, M. Kaul, K.-R. Mueller, S. Nakajima, ''Sharing Hash Codes for Multiple Purposes,''
Japanese Journal of Statistics and Data Science, vol.1(1), pp.215-246, 2018.

- We will present the following paper in ICCV-RSL workshop:

S. Dogadov, A. Masegosa, S. Nakajima, ''Variational Robust Subspace Clustering with Mean Update Algorithm''
ICCV Workshop on Robust Subspace Learning and Applications in Computer Vision (RSL-CV), (Venice, Italy, October 28, 2017).

- The following papers have been accepted:

N. Görnitz, L. A. Lima, K.-R. Müller, M. Kloft, S. Nakajima, ''Support Vector Data Descriptions and K-means Clustering: One Class?''
IEEE Transactions on Neural Networks and Learning Systems, vol.29(9), pp.3994-4006, 2018.

W. Samek, S. Nakajima, M. Kawanabe, K.-R. Müller, ''On Robust Parameter Estimation in Brain-Computer Interfacing,''
Journal of Neural Engineering, vol.14, 061001, 2017.

- Our paper has been accepted for the journal track of a conference:

S. Mandt, F. Wenzel, S. Nakajima, J. Cunningham, C. Lippert, M. Kloft, ''Sparse Probit Linear Mixed Model,''
ECML/PKDD2017 (Skopje, Macedonia, Semptember 18-22, 2017),
Machine Learning, vol.106, pp.1621-1642 (ECMLPKDD Special Issue), 2017.

- Our conference papers on graph analysis has been accepted:

J. Höner, S. Nakajima, A. Bauer, K.-R. Müller, N. Görnitz, ''Minimizing Trust Leaks for Robust Sybil Detection,''
ICML2017, (Sydney Australia, August 6-11, 2017).

- Our journal papers on structural learning have been accepted:

L. A. Lima, N. Görnitz, L. E. Varella, M. Vellascob, K.-R. Müller, S. Nakajima, ''Porosity Estimation by Semi-supervised Learning with Sparsely Available Labeled Samples,''
Computers and Geosciences, vol.106, pp.33-48, 2017.

N. Görnitz, L. A. Lima, L. E. Varella, K.-R. Müller, S. Nakajima, ''Transductive Regression for Data with Latent Dependency Structure,''
IEEE Transactions on Neural Networks and Learning Systems, vol.29(7), pp.2743-2756, 2018.

A. Bauer, S. Nakajima, K.-R. Müller, ''Efficient Exact Inference with Loss Augmented Objective in Structured Learning,''
IEEE Transactions on Neural Networks and Learning Systems, vol.28, pp.2566-2579, 2017.

- Our journal papers have been accepted:

K. Nagata, J. Kitazono, S. Nakajima, S. Eifuku, R. Tamura, M. Okada, ''An Exhaustive Search and Stability of Sparse Estimation for Feature Selection Problem,''
IPSJ Transactions on Mathematical Modeling and Its Applications, vol.8, pp.25-32, 2015.

S. Nakajima, R. Tomioka, M. Sugiyama, S. D. Babacan, ''Condition for Perfect Dimensionality Recovery by Variational Bayesian PCA,''
Journal of Machine Learning Research, vol.16, pp.3757-3811, 2015, Matlab/Python code is now available from here!.

- The following papers have been accepted:

S. Nakajima, I. Sato, M. Sugiyama, K. Watanabe, H. Kobayashi, '' Analysis of Variational Bayesian Latent Dirichlet Allocation: Weaker Sparsity than MAP,''

NIPS2014, (Montreal, Canada, December 8-13, 2014).

S. D. Babacan, S. Nakajima, M. N. Do, ''Bayesian Group-Sparse modeling and Variational Inference,''
IEEE Transactions on Signal Processing, vol.62, pp.2906-2921, 2014.

S. Nakajima, M. Sugiyama, ''Analysis of Empirical MAP and Empirical Partially Bayes: Can They be Alternatives to Variational Bayes?''
AISTATS2014, (Reykjavik, Iceland, April 22-25, 2014),
Selected for Notable Paper Award.

Our conference papers have been accepted:

S. Nakajima, A. Takeda, S. D. Babacan, M. Sugiyama, I. Takeuchi, ''Global Solver and Its Efficient Approximation for Variational Bayesian Low-Rank Subspace Clustering,''
NIPS2013, (Lake Tahoe, USA, December 5-10, 2013).

I. Takeuchi, T. Hongo, M. Sugiyama, S. Nakajima, ''Parametric Task Learning,''
NIPS2013, (Lake Tahoe, USA, December 5-10, 2013).

Our conference paper on light field acquisition has been accepted:

P. Ruiz, J. Mateos, M. C. Cardenas, S. Nakajima, R. Molina, A. K. Katsaggelos, ''Light Field Acquisition From Blurred Observations Using a Programmable Coded Aperture Camera,''
EUSIPCO2013, (Marrakech, Morocco, September 9-13, 2013).

- Our journal papers have been accepted:

S. Nakajima, M. Sugiyama, S. D. Babacan, ''Variational Bayesian Sparse Additive Matrix Factorization,''
Machine Learning, vol.92, pp.319-347, DOI:10.1007/s10994-013-5347-6,
(Special Issue of Selected Papers of ACML 2012), 2013.

S. Nakajima, M. Sugiyama, S. D. Babacan, R. Tomioka, ''Global Analytic Solution of Fully-observed Variational Bayesian Matrix Factorization,''
Journal of Machine Learning Research, vol.14, pp.1-37, 2013.

- Our conference papers on variational Bayesian methods have been accepted:

S. Nakajima, R. Tomioka, M. Sugiyama, S. D. Babacan, ''Perfect Dimensionality Recovery by Variational Bayesian PCA,''
NIPS2012, (Lake Tahoe, USA, December 3-8, 2012), poster.

S. D. Babacan, S. Nakajima, M. Do, ''Probabilistic Low-Rank Subspace Clustering,''
NIPS2012, (Lake Tahoe, USA, December 3-8, 2012).

S. Nakajima, M. Sugiyama, S. D. Babacan, ''Sparse Additive Matrix Factorization for Robust PCA and Its Generalization,''
ACML2012, (Singapore, November 4-6, 2012), slides.

- Our journal paper on multiple kernel learning has been accepted:

A. Binder, S. Nakajima, M. Kloft, C. Mueller, W. Samek, U. Brefeld, K.-R. Mueller, M. Kawanabe, ''Insights from Classifying Visual Concepts with Multiple Kernel Learning,''
PLoS ONE, 7(8): e38897.doi:10.1371/journal.pone.0038897, 2012. 

Prof. David MacKay's view of evidence. Evidence-based model selection prefers simpler models. This is not accurate for non-identifiable models unless Jeffreys prior is adopted.

A more accurate view for non-identifiable models. Simpler models are prefered even without model selection, due to density non-uniformity in the parameter space.
Research Activities

    * Image Analysis

    * Machine Learning for Photolithography
      Steppers/scanners are ones of Nikon's most important products. I have worked for several years on alignment --- to align a silicon wafer to a mask in a few nano meters of accuracy. I developed a series of signal prosessing algorithms,

        S. Nakajima, Y. Kanaya, N. Magome, ''Improving the Measurement Algorithm for Alignment,''
        SPIE Microlithography 2001 (Santa Clara, U.S.A., March, 2001),

      applied a model selection procedure to a wafer distortion model,

        S. Nakajima, S. Watanabe,
        ''Simulation Data Generation from Extended EGA Model and Optimization of Alignment Strategy for Lithography,''
        ISITA2004 (Parma, Italy, October 10-13, 2004),

      and utilized mixture models for outlier rejection.

        S. Nakajima, Y. Kanaya, M. Li, T. Sugihara, A. Sukegawa, N. Magome, ''Outlier Rejection with Mixture Models in Alignment,''
        SPIE Microlithography 2003 (Santa Clara, U.S.A., March, 2003).

      Recently, Prof. Masashi Sugiyama and I have applied a novel active learning method, called P-ALICE, to finding the best configuration of sampling positions on a wafer, at which images of alignemnt marks are captured at the expense of throughput.

Subpages (1): Publication List