IEICE Trans

Masashi SUGIYAMA

Learning from Noisy Complementary Labels with Robust Loss Functions Hiroki ISHIGURO Takashi ISHIDA Masashi SUGIYAMA
Publication: Publication Date: 2022/02/01 Vol. E105-D No. 2 pp. 364-376 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: complementary label, label noise, robust loss function, loss correction,
	Summary \| Full Text:PDF

Stochastic Divergence Minimization for Biterm Topic Models Zhenghang CUI Issei SATO Masashi SUGIYAMA
Publication: Publication Date: 2018/03/01 Vol. E101-D No. 3 pp. 668-677 Type of Manuscript: PAPER Category: Data Engineering, Web Information Systems Keyword: short text, topic model, biterm, stochastic inference algorithm,
	Summary \| Full Text:PDF

Computationally Efficient Class-Prior Estimation under Class Balance Change Using Energy Distance Hideko KAWAKUBO Marthinus Christoffel DU PLESSIS Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2016/01/01 Vol. E99-D No. 1 pp. 176-186 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: class balance change, class-prior estimation, energy distance,
	Summary \| Full Text:PDF

Direct Density Ratio Estimation with Convolutional Neural Networks with Application in Outlier Detection Hyunha NAM Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2015/05/01 Vol. E98-D No. 5 pp. 1073-1079 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: density ratio estimation, convolutional neural network, outlier detection,
	Summary \| Full Text:PDF

Unsupervised Dimension Reduction via Least-Squares Quadratic Mutual Information Janya SAINUI Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2014/10/01 Vol. E97-D No. 10 pp. 2806-2809 Type of Manuscript: LETTER Category: Artificial Intelligence, Data Mining Keyword: unsupervised dimension reduction, quadratic mutual information, least-squares density difference, Epanechnikov kernel, hyperparameter tuning,
	Summary \| Full Text:PDF

Constrained Least-Squares Density-Difference Estimation Tuan Duong NGUYEN Marthinus Christoffel DU PLESSIS Takafumi KANAMORI Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2014/07/01 Vol. E97-D No. 7 pp. 1822-1829 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: density difference, asymptotic variance, L²-distance, bias, class-balance change, two-sample homogeneity test,
	Summary \| Full Text:PDF

Tree-Based Ensemble Multi-Task Learning Method for Classification and Regression Jaak SIMM Ildefons MAGRANS DE ABRIL Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2014/06/01 Vol. E97-D No. 6 pp. 1677-1681 Type of Manuscript: LETTER Category: Pattern Recognition Keyword: machine learning, multi-task learning, tree-based methods, ensemble methods,
	Summary \| Full Text:PDF

Class Prior Estimation from Positive and Unlabeled Data Marthinus Christoffel DU PLESSIS Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2014/05/01 Vol. E97-D No. 5 pp. 1358-1362 Type of Manuscript: LETTER Category: Artificial Intelligence, Data Mining Keyword: class-prior change, outlier detection, positive and unlabeled learning, divergence estimation, pearson divergence,
	Summary \| Full Text:PDF

Computationally Efficient Estimation of Squared-Loss Mutual Information with Multiplicative Kernel Models Tomoya SAKAI Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2014/04/01 Vol. E97-D No. 4 pp. 968-971 Type of Manuscript: LETTER Category: Fundamentals of Information Systems Keyword: squared-loss mutual information, least-squares mutual information, density ratio estimation, multiplicative kernel models, independence test,
	Summary \| Full Text:PDF

Direct Approximation of Quadratic Mutual Information and Its Application to Dependence-Maximization Clustering Janya SAINUI Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2013/10/01 Vol. E96-D No. 10 pp. 2282-2285 Type of Manuscript: LETTER Category: Artificial Intelligence, Data Mining Keyword: quadratic mutual information, least-square density-difference estimation, information theoretic clustering,
	Summary \| Full Text:PDF

Computationally Efficient Multi-Label Classification by Least-Squares Probabilistic Classifiers Hyunha NAM Hirotaka HACHIYA Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2013/08/01 Vol. E96-D No. 8 pp. 1871-1874 Type of Manuscript: LETTER Category: Fundamentals of Information Systems Keyword: multi-label classification, least-squares probabilistic classifier,
	Summary \| Full Text:PDF

Feature Selection via ₁-Penalized Squared-Loss Mutual Information Wittawat JITKRITTUM Hirotaka HACHIYA Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2013/07/01 Vol. E96-D No. 7 pp. 1513-1524 Type of Manuscript: PAPER Category: Pattern Recognition Keyword: feature selection, ₁-regularization, squared-loss mutual information, density-ratio estimation, dimensionality reduction,
	Summary \| Full Text:PDF

Artist Agent: A Reinforcement Learning Approach to Automatic Stroke Generation in Oriental Ink Painting Ning XIE Hirotaka HACHIYA Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2013/05/01 Vol. E96-D No. 5 pp. 1134-1144 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: painterly rendering, stroke-based rendering, reinforcement learning, policy gradient,
	Summary \| Full Text:PDF

Winning the Kaggle Algorithmic Trading Challenge with the Composition of Many Models and Feature Engineering Ildefons MAGRANS DE ABRIL Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2013/03/01 Vol. E96-D No. 3 pp. 742-745 Type of Manuscript: LETTER Category: Artificial Intelligence, Data Mining Keyword: Kaggle challenge, model architecture, boosting, feature selection, high frequency trading, liquidity shock, maximal information coefficient,
	Summary \| Full Text:PDF

Multi-Task Approach to Reinforcement Learning for Factored-State Markov Decision Problems Jaak SIMM Masashi SUGIYAMA Hirotaka HACHIYA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2012/10/01 Vol. E95-D No. 10 pp. 2426-2437 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: reinforcement learning, multi-task learning, transfer learning, factored state models, Markov decision process,
	Summary \| Full Text:PDF

On Kernel Parameter Selection in Hilbert-Schmidt Independence Criterion Masashi SUGIYAMA Makoto YAMADA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2012/10/01 Vol. E95-D No. 10 pp. 2564-2567 Type of Manuscript: LETTER Category: Artificial Intelligence, Data Mining Keyword: Hilbert-Schmidt independence criterion, least-squares mutual information, cross-validation, Gaussian kernel,
	Summary \| Full Text:PDF

Early Stopping Heuristics in Pool-Based Incremental Active Learning for Least-Squares Probabilistic Classifier Tsubasa KOBAYASHI Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2012/08/01 Vol. E95-D No. 8 pp. 2065-2073 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: pool-based incremental active learning, early stopping, least-squares probabilistic classifier, semi-supervised learning,
	Summary \| Full Text:PDF

FOREWORD Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2011/10/01 Vol. E94-D No. 10 pp. 1845-1845 Type of Manuscript: FOREWORD Category: Keyword:
	Summary \| Full Text:PDF

Least-Squares Independence Test Masashi SUGIYAMA Taiji SUZUKI
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2011/06/01 Vol. E94-D No. 6 pp. 1333-1336 Type of Manuscript: LETTER Category: Artificial Intelligence, Data Mining Keyword: independence test, density ratio estimation, unconstrained least-squares importance fitting, squared-loss mutual information,
	Summary \| Full Text:PDF

Improving the Accuracy of Least-Squares Probabilistic Classifiers Makoto YAMADA Masashi SUGIYAMA Gordon WICHERN Jaak SIMM
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2011/06/01 Vol. E94-D No. 6 pp. 1337-1340 Type of Manuscript: LETTER Category: Pattern Recognition Keyword: least-squares probabilistic classifier, kernel logistic regression, density ratio, PASCAL VOC 2010, freesound,
	Summary \| Full Text:PDF

Lighting Condition Adaptation for Perceived Age Estimation Kazuya UEKI Masashi SUGIYAMA Yasuyuki IHARA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2011/02/01 Vol. E94-D No. 2 pp. 392-395 Type of Manuscript: LETTER Category: Image Recognition, Computer Vision Keyword: face recognition, age estimation, covariate shift adaptation, lighting condition change, Kullback-Leibler importance estimation procedure, importance-weighted regularized least-squares,
	Summary \| Full Text:PDF

Superfast-Trainable Multi-Class Probabilistic Classifier by Least-Squares Posterior Fitting Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/10/01 Vol. E93-D No. 10 pp. 2690-2701 Type of Manuscript: Special Section PAPER (Special Section on Data Mining and Statistical Science) Category: Keyword: probabilistic classification, kernel logistic regression, class-posterior probability, squared-loss,
	Summary \| Full Text:PDF

Direct Importance Estimation with a Mixture of Probabilistic Principal Component Analyzers Makoto YAMADA Masashi SUGIYAMA Gordon WICHERN Jaak SIMM
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/10/01 Vol. E93-D No. 10 pp. 2846-2849 Type of Manuscript: LETTER Category: Fundamentals of Information Systems Keyword: importance, Kullback-Leibler importance estimation procedure, mixture of probabilistic principal component analyzers, expectation-maximization algorithm,
	Summary \| Full Text:PDF

FOREWORD Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/10/01 Vol. E93-D No. 10 pp. 2671-2671 Type of Manuscript: FOREWORD Category: Keyword:
	Summary \| Full Text:PDF

A Semi-Supervised Approach to Perceived Age Prediction from Face Images Kazuya UEKI Masashi SUGIYAMA Yasuyuki IHARA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/10/01 Vol. E93-D No. 10 pp. 2875-2878 Type of Manuscript: LETTER Category: Image Recognition, Computer Vision Keyword: perceived age estimation, active learning, weighted regression, semi-supervised learning, manifold regularization, human age perception,
	Summary \| Full Text:PDF

Least Absolute Policy Iteration--A Robust Approach to Value Function Approximation Masashi SUGIYAMA Hirotaka HACHIYA Hisashi KASHIMA Tetsuro MORIMURA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/09/01 Vol. E93-D No. 9 pp. 2555-2565 Type of Manuscript: PAPER Category: Artificial Intelligence, Data Mining Keyword: reinforcement learning, value function approximation, least-squares policy iteration, outlier, ₁-loss function, linear programming,
	Summary \| Full Text:PDF

Spectral Methods for Thesaurus Construction Nobuyuki SHIMIZU Masashi SUGIYAMA Hiroshi NAKAGAWA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/06/01 Vol. E93-D No. 6 pp. 1378-1385 Type of Manuscript: Special Section PAPER (Special Section on Info-Plosion) Category: Natural Language Processing Keyword: synonym acquisition, synonym extraction, thesaurus, spectral clustering, graph laplacian,
	Summary \| Full Text:PDF

Theoretical Analysis of Density Ratio Estimation Takafumi KANAMORI Taiji SUZUKI Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Publication Date: 2010/04/01 Vol. E93-A No. 4 pp. 787-798 Type of Manuscript: PAPER Category: Algorithms and Data Structures Keyword: density ratio estimation, density estimation, logistic regression, asymptotic analysis, Gaussian assumption,
	Summary \| Full Text:PDF

Least-Squares Conditional Density Estimation Masashi SUGIYAMA Ichiro TAKEUCHI Taiji SUZUKI Takafumi KANAMORI Hirotaka HACHIYA Daisuke OKANOHARA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2010/03/01 Vol. E93-D No. 3 pp. 583-594 Type of Manuscript: PAPER Category: Pattern Recognition Keyword: conditional density estimation, multimodality, heteroscedastic noise, direct density ratio estimation, transition estimation,
	Summary \| Full Text:PDF

Direct Importance Estimation with Gaussian Mixture Models Makoto YAMADA Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2009/10/01 Vol. E92-D No. 10 pp. 2159-2162 Type of Manuscript: LETTER Category: Pattern Recognition Keyword: importance weight, KLIEP, Gaussian mixture models, EM algorithm,
	Summary \| Full Text:PDF

Recent Advances and Trends in Large-Scale Kernel Methods Hisashi KASHIMA Tsuyoshi IDE Tsuyoshi KATO Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2009/07/01 Vol. E92-D No. 7 pp. 1338-1353 Type of Manuscript: INVITED PAPER (Special Section on Large Scale Algorithms for Learning and Optimization) Category: Keyword: kernel methods, support vector machines, kernel trick, low-rank approximation, optimization, structured data,
	Summary \| Full Text:PDF

On Computational Issues of Semi-Supervised Local Fisher Discriminant Analysis Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2009/05/01 Vol. E92-D No. 5 pp. 1204-1208 Type of Manuscript: LETTER Category: Artificial Intelligence and Cognitive Science Keyword: dimensionality reduction, semi-supervised learning, semi-supervised local Fisher discriminant analysis, sparsity, generalized eigenvalue problem,
	Summary \| Full Text:PDF

Approximating the Best Linear Unbiased Estimator of Non-Gaussian Signals with Gaussian Noise Masashi SUGIYAMA Motoaki KAWANABE Gilles BLANCHARD Klaus-Robert MULLER
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2008/05/01 Vol. E91-D No. 5 pp. 1577-1580 Type of Manuscript: LETTER Category: Pattern Recognition Keyword: signal denoising, best linear unbiased estimator (BLUE), non-Gaussian component analysis (NGCA), Gaussian noise,
	Summary \| Full Text:PDF

A New Meta-Criterion for Regularized Subspace Information Criterion Yasushi HIDAKA Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2007/11/01 Vol. E90-D No. 11 pp. 1779-1786 Type of Manuscript: PAPER Category: Pattern Recognition Keyword: supervised learning, generalization capability, model selection, unbiased estimator, regularized subspace information criterion,
	Summary \| Full Text:PDF

Analytic Optimization of Adaptive Ridge Parameters Based on Regularized Subspace Information Criterion Shun GOKITA Masashi SUGIYAMA Keisuke SAKURAI
Publication: IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Publication Date: 2007/11/01 Vol. E90-A No. 11 pp. 2584-2592 Type of Manuscript: PAPER Category: Neural Networks and Bioengineering Keyword: supervised learning, generalization error, model selection, regularized subspace information criterion,
	Summary \| Full Text:PDF

Generalization Error Estimation for Non-linear Learning Methods Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Publication Date: 2007/07/01 Vol. E90-A No. 7 pp. 1496-1499 Type of Manuscript: LETTER Category: Neural Networks and Bioengineering Keyword: supervised learning, generalization capability, model selection, subspace information criterion, non-linear learning,
	Summary \| Full Text:PDF

Analytic Optimization of Shrinkage Parameters Based on Regularized Subspace Information Criterion Masashi SUGIYAMA Keisuke SAKURAI
Publication: IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Publication Date: 2006/08/01 Vol. E89-A No. 8 pp. 2216-2225 Type of Manuscript: PAPER Category: Neural Networks and Bioengineering Keyword: supervised learning, generalization capability, model selection, shrinkage estimator, regularized subspace information criterion,
	Summary \| Full Text:PDF

Constructing Kernel Functions for Binary Regression Masashi SUGIYAMA Hidemitsu OGAWA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2006/07/01 Vol. E89-D No. 7 pp. 2243-2249 Type of Manuscript: PAPER Category: Pattern Recognition Keyword: supervised learning, regression, kernel methods, kernel functions, Karhunen-Loeve expansion, principal component analysis, binary regression, Gaussian kernel,
	Summary \| Full Text:PDF

Active Learning with Model Selection -- Simultaneous Optimization of Sample Points and Models for Trigonometric Polynomial Models Masashi SUGIYAMA Hidemitsu OGAWA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2003/12/01 Vol. E86-D No. 12 pp. 2753-2763 Type of Manuscript: PAPER Category: Pattern Recognition Keyword: supervised learning, generalization capability, active learning, model selection, trigonometric polynomial space,
	Summary \| Full Text:PDF

Improving Precision of the Subspace Information Criterion Masashi SUGIYAMA
Publication: IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Publication Date: 2003/07/01 Vol. E86-A No. 7 pp. 1885-1895 Type of Manuscript: PAPER Category: Neural Networks and Bioengineering Keyword: machine learning, generalization capability, subspace information criterion, generalized inverse, unbiased estimator, variance,
	Summary \| Full Text:PDF

Incremental Construction of Projection Generalizing Neural Networks Masashi SUGIYAMA Hidemitsu OGAWA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2002/09/01 Vol. E85-D No. 9 pp. 1433-1442 Type of Manuscript: PAPER Category: Biocybernetics, Neurocomputing Keyword: generalization capability, incremental learning, pseudo biorthogonal basis (PBOB), projection learning, projection generalizing neural network,
	Summary \| Full Text:PDF

Active Learning for Optimal Generalization in Trigonometric Polynomial Models Masashi SUGIYAMA Hidemitsu OGAWA
Publication: IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Publication Date: 2001/09/01 Vol. E84-A No. 9 pp. 2319-2329 Type of Manuscript: PAPER Category: Algorithms and Data Structures Keyword: supervised learning, active learning, generalization capability, trigonometric polynomial space, pseudo orthogonal bases,
	Summary \| Full Text:PDF

Subspace Information Criterion for Image Restoration--Optimizing Parameters in Linear Filters Masashi SUGIYAMA Daisuke IMAIZUMI Hidemitsu OGAWA
Publication: IEICE TRANSACTIONS on Information and Systems Publication Date: 2001/09/01 Vol. E84-D No. 9 pp. 1249-1256 Type of Manuscript: PAPER Category: Image Processing, Image Pattern Recognition Keyword: image restoration, mean squared error, subspace information criterion, moving-average filter, model selection,
	Summary \| Full Text:PDF