no code implementations • ICML 2020 • Futoshi Futami, Issei Sato, Masashi Sugiyama
Compared with the naive parallel-chain SGLD that updates multiple particles independently, ensemble methods update particles with their interactions.
1 code implementation • 14 Feb 2024 • Keitaro Sakamoto, Issei Sato
End-to-end (E2E) training, optimizing the entire model through error backpropagation, fundamentally supports the advancements of deep learning.
no code implementations • 27 Oct 2023 • Shuo Wang, Issei Sato
Furthermore, we show that the existing parameter saliency method exhibits a bias against the depth of layers in deep neural networks.
no code implementations • 10 Oct 2023 • Takeshi Koshizuka, Masahiro Fujisawa, Yusuke Tanaka, Issei Sato
Building upon this observation, we also propose an edge of chaos initialization scheme for FNO to mitigate the negative initialization bias leading to training instability.
no code implementations • 26 Jul 2023 • Tokio Kajitsuka, Issei Sato
Existing analyses of the expressive capacity of Transformer models have required excessively deep layers for data memorization, leading to a discrepancy with the Transformers actually used in practice.
no code implementations • 26 May 2023 • Naoya Hasegawa, Issei Sato
Recognition problems in long-tailed data, in which the sample size per class is heavily skewed, have gained importance because the distribution of the sample size per class in a dataset is generally exponential unless the sample size is intentionally adjusted.
no code implementations • 2 Jun 2022 • Futoshi Futami, Tomoharu Iwata, Naonori Ueda, Issei Sato, Masashi Sugiyama
Bayesian deep learning plays an important role especially for its ability evaluating epistemic uncertainty (EU).
no code implementations • 15 May 2022 • Keitaro Sakamoto, Issei Sato
The lottery ticket hypothesis (LTH) has attracted attention because it can explain why over-parameterized models often show high generalization ability.
no code implementations • 29 Apr 2022 • Takahiro Suzuki, Shouhei Hanaoka, Issei Sato
An obstacle in developing a CAD system for a disease is that the number of medical images is typically too small to improve the performance of the machine learning model.
no code implementations • 18 Apr 2022 • Kento Nozawa, Issei Sato
Representation learning enables us to automatically extract generic feature representations from a dataset to solve another machine learning task.
1 code implementation • 11 Apr 2022 • Takeshi Koshizuka, Issei Sato
Population dynamics is the study of temporal and spatial variation in the size of populations of organisms and is a major part of population ecology.
no code implementations • 11 Oct 2021 • Mingcheng Hou, Issei Sato
The prototypical network is a prototype classifier based on meta-learning and is widely used for few-shot learning because it classifies unseen examples by constructing class-specific prototypes without adjusting hyper-parameters during meta-testing.
no code implementations • 29 Sep 2021 • Zeke Xie, Xinrui Wang, Huishuai Zhang, Issei Sato, Masashi Sugiyama
Specifically, we disentangle the effects of Adaptive Learning Rate and Momentum of the Adam dynamics on saddle-point escaping and flat minima selection.
no code implementations • ICLR 2022 • Seiya Tokui, Issei Sato
We propose a framework to analyze how multivariate representations disentangle ground-truth generative factors.
no code implementations • NeurIPS 2021 • Futoshi Futami, Tomoharu Iwata, Naonori Ueda, Issei Sato, Masashi Sugiyama
First, we provide a new second-order Jensen inequality, which has the repulsion term based on the loss function.
1 code implementation • 17 Mar 2021 • Naoki Kobayashi, Taro Sekiyama, Issei Sato, Hiroshi Unno
Another application is to a new program development framework called oracle-based programming, which is a neural-network-guided variation of Solar-Lezama's program synthesis by sketching.
no code implementations • 24 Feb 2021 • Kenshin Abe, Takanori Maehara, Issei Sato
We study the problem of modeling a binary operation that satisfies some algebraic requirements.
1 code implementation • NeurIPS 2021 • Kento Nozawa, Issei Sato
Instance discriminative self-supervised representation learning has been attracted attention thanks to its unsupervised nature and informative feature representation for downstream tasks.
1 code implementation • 1 Feb 2021 • Nan Lu, Shida Lei, Gang Niu, Issei Sato, Masashi Sugiyama
SSC can be solved by a standard (multi-class) classification method, and we use the SSC solution to obtain the final binary classifier through a certain linear-fractional transformation.
1 code implementation • NeurIPS 2023 • Zeke Xie, Zhiqiang Xu, Jingzhao Zhang, Issei Sato, Masashi Sugiyama
Weight decay is a simple yet powerful regularization technique that has been very widely used in training of deep neural networks (DNNs).
1 code implementation • 12 Nov 2020 • Zeke Xie, Fengxiang He, Shaopeng Fu, Issei Sato, DaCheng Tao, Masashi Sugiyama
Thus it motivates us to design a similar mechanism named {\it artificial neural variability} (ANV), which helps artificial neural networks learn some advantages from ``natural'' neural networks.
no code implementations • 28 Sep 2020 • Zeke Xie, Issei Sato, Masashi Sugiyama
\citet{loshchilov2018decoupled} demonstrated that $L_{2}$ regularization is not identical to weight decay for adaptive gradient methods, such as Adaptive Momentum Estimation (Adam), and proposed Adam with Decoupled Weight Decay (AdamW).
1 code implementation • 3 Aug 2020 • Zhenghang Cui, Issei Sato
We then propose an efficient adaptive labeling algorithm using the proposed oracle and the positivity comparison oracle.
no code implementations • 3 Jul 2020 • Takahiro Mimori, Keiko Sasada, Hirotaka Matsui, Issei Sato
We propose an evaluation framework for class probability estimates (CPEs) in the presence of label uncertainty, which is commonly observed as diagnosis disagreement between experts in the medical domain.
1 code implementation • 29 Jun 2020 • Zeke Xie, Xinrui Wang, Huishuai Zhang, Issei Sato, Masashi Sugiyama
Specifically, we disentangle the effects of Adaptive Learning Rate and Momentum of the Adam dynamics on saddle-point escaping and minima selection.
no code implementations • 15 Jun 2020 • Kei Mukaiyama, Issei Sato, Masashi Sugiyama
The prototypical network (ProtoNet) is a few-shot learning framework that performs metric learning and classification using the distance to prototype representations of each class.
no code implementations • 13 Jun 2020 • Masahiro Fujisawa, Takeshi Teshima, Issei Sato, Masashi Sugiyama
Approximate Bayesian computation (ABC) is a likelihood-free inference method that has been employed in various applications.
no code implementations • 11 Jun 2020 • Han Bao, Takuya Shimada, Liyuan Xu, Issei Sato, Masashi Sugiyama
A classifier built upon the representations is expected to perform well in downstream classification; however, little theory has been given in literature so far and thereby the relationship between similarity and classification has remained elusive.
no code implementations • 8 May 2020 • Yuki Koyama, Issei Sato, Masataka Goto
To help users respond to plane-search queries, we also propose using a gallery-based interface that provides options in the two-dimensional subspace arranged in an adaptive grid view.
no code implementations • 10 Mar 2020 • Hideaki Imamura, Nontawat Charoenphakdee, Futoshi Futami, Issei Sato, Junya Honda, Masashi Sugiyama
If the black-box function varies with time, then time-varying Bayesian optimization is a promising framework.
no code implementations • ICLR 2021 • Zeke Xie, Issei Sato, Masashi Sugiyama
Stochastic Gradient Descent (SGD) and its variants are mainstream methods for training deep networks in practice.
1 code implementation • ICML 2020 • Takeshi Teshima, Issei Sato, Masashi Sugiyama
We take the structural equations in causal modeling as an example and propose a novel DA method, which is shown to be useful both theoretically and experimentally.
no code implementations • 20 Nov 2019 • Soma Yokoi, Issei Sato
The current interpretation of stochastic gradient descent (SGD) as a stochastic process lacks generality in that its numerical scheme restricts continuous-time dynamics as well as the loss function and the distribution of gradient noise.
1 code implementation • 24 Jul 2019 • Zhenghang Cui, Nontawat Charoenphakdee, Issei Sato, Masashi Sugiyama
Although learning from triplet comparison data has been considered in many applications, an important fundamental question of whether we can learn a classifier only from triplet comparison data has remained unanswered.
no code implementations • 24 Jun 2019 • Toby Chong Long Hin, I-Chao Shen, Issei Sato, Takeo Igarashi
We present a human-in-the-optimization method that allows users to directly explore and search the latent vector space of generative image modeling.
2 code implementations • 28 May 2019 • Kenshin Abe, Zijian Xu, Issei Sato, Masashi Sugiyama
There have been increasing challenges to solve combinatorial optimization problems by machine learning.
no code implementations • 2 May 2019 • Xi Yang, Bojian Wu, Issei Sato, Takeo Igarashi
Deep neural networks (DNNs) have a high accuracy on image classification tasks.
no code implementations • 26 Apr 2019 • Takuya Shimada, Han Bao, Issei Sato, Masashi Sugiyama
In this paper, we derive an unbiased risk estimator which can handle all of similarities/dissimilarities and unlabeled data.
no code implementations • 22 Mar 2019 • Hiroaki Adachi, Yoko Kawamura, Keiji Nakagawa, Ryoichi Horisaki, Issei Sato, Satoko Yamaguchi, Katsuhito Fujiu, Kayo Waki, Hiroyuki Noji, Sadao Ota
Imaging flow cytometry shows significant potential for increasing our understanding of heterogeneous and complex life systems and is useful for biomedical applications.
no code implementations • 14 Mar 2019 • Issei Sato
Computational ghost imaging is an imaging technique in which an object is imaged from light collected using a single-pixel detector with no spatial resolution.
no code implementations • 7 Mar 2019 • Soma Yokoi, Takuma Otsuka, Issei Sato
Although SGLD is designed for unbounded random variables, many practical models incorporate variables with boundaries such as non-negative ones or those in a finite interval.
no code implementations • 12 Feb 2019 • Kento Nozawa, Issei Sato
Learning sentence vectors from an unlabeled corpus has attracted attention because such vectors can represent sentences in a lower dimensional and continuous space.
no code implementations • 4 Feb 2019 • Takuo Kaneko, Issei Sato, Masashi Sugiyama
We consider the problem of online multiclass classification with partial feedback, where an algorithm predicts a class for a new instance in each round and only receives its correctness.
no code implementations • 1 Feb 2019 • Masahiro Fujisawa, Issei Sato
We theoretically show that, with our method, the variance of the gradient estimator decreases as optimization proceeds and that a learning rate scheduler function helps improve the convergence.
no code implementations • 31 Jan 2019 • Taira Tsuchiya, Nontawat Charoenphakdee, Issei Sato, Masashi Sugiyama
We further provide an estimation error bound to show that our risk estimator is consistent.
no code implementations • ICML 2020 • Yusuke Tsuzuku, Issei Sato, Masashi Sugiyama
However, existing definitions of the flatness are known to be sensitive to the rescaling of parameters.
no code implementations • 13 Sep 2018 • Takeshi Teshima, Miao Xu, Issei Sato, Masashi Sugiyama
On the other hand, matrix completion (MC) methods can recover a low-rank matrix from various information deficits by using the principle of low-rank completion.
no code implementations • CVPR 2019 • Yusuke Tsuzuku, Issei Sato
Data-agnostic quasi-imperceptible perturbations on inputs are known to degrade recognition accuracy of deep convolutional networks severely.
no code implementations • 11 Sep 2018 • Seiichi Kuroki, Nontawat Charoenphakdee, Han Bao, Junya Honda, Issei Sato, Masashi Sugiyama
A previously proposed discrepancy that does not use the source domain labels requires high computational cost to estimate and may lead to a loose generalization error bound in the target domain.
no code implementations • 21 May 2018 • Futoshi Futami, Zhenghang Cui, Issei Sato, Masashi Sugiyama
Another example is the Stein points (SP) method, which minimizes kernelized Stein discrepancy directly.
no code implementations • 12 Mar 2018 • Hongyi Ding, Young Lee, Issei Sato, Masashi Sugiyama
We present the first framework for Gaussian-process-modulated Poisson processes when the temporal data appear in the form of panel counts.
1 code implementation • ICML 2018 • Hideaki Imamura, Issei Sato, Masashi Sugiyama
In this paper, we derive a minimax error rate under more practical setting for a broader class of crowdsourcing models including the DS model as a special case.
no code implementations • 12 Feb 2018 • Ryosuke Kamesawa, Issei Sato, Masashi Sugiyama
A state-of-the-art method of Gaussian process classification (GPC) with privileged information is GPC+, which incorporates privileged information into a noise term of the likelihood.
2 code implementations • NeurIPS 2018 • Yusuke Tsuzuku, Issei Sato, Masashi Sugiyama
High sensitivity of neural networks against malicious perturbations on inputs causes security concerns.
no code implementations • 22 Nov 2017 • Zeke Xie, Issei Sato
The contribution of this work is two-fold, a novel ensemble regression algorithm inspired by quantum mechanics and the theoretical connection between quantum interpretations and machine learning algorithms.
1 code implementation • 18 Oct 2017 • Futoshi Futami, Issei Sato, Masashi Sugiyama
In this paper, based on Zellner's optimization and variational formulation of Bayesian inference, we propose an outlier-robust pseudo-Bayesian variational method by replacing the Kullback-Leibler divergence used for data fitting to a robust divergence such as the beta- and gamma-divergences.
no code implementations • NeurIPS 2017 • Iku Ohama, Issei Sato, Takuya Kida, Hiroki Arimura
In order to ensure that the model shrinkage effect of the EPM works in an appropriate manner, we proposed two novel generative constructions of the EPM: CEPM incorporating constrained gamma priors, and DEPM incorporating Dirichlet priors instead of the gamma priors.
no code implementations • ICML 2017 • Seiya Tokui, Issei Sato
The framework gives a natural derivation of the optimal estimator that can be interpreted as a special case of the likelihood-ratio method so that we can evaluate the optimal degree of practical techniques with it.
no code implementations • NeurIPS 2017 • Futoshi Futami, Issei Sato, Masashi Sugiyama
Exponential family distributions are highly useful in machine learning since their calculation can be performed efficiently through natural parameters.
1 code implementation • 19 May 2017 • Hongyi Ding, Mohammad Emtiyaz Khan, Issei Sato, Masashi Sugiyama
We model the intensity of each sequence as an infinite mixture of latent functions, each of which is obtained using a function drawn from a Gaussian process.
no code implementations • 1 May 2017 • Zhenghang Cui, Issei Sato, Masashi Sugiyama
As the emergence and the thriving development of social networks, a huge number of short texts are accumulated and need to be processed.
1 code implementation • 22 Apr 2017 • Han Bao, Tomoya Sakai, Issei Sato, Masashi Sugiyama
Multiple instance learning (MIL) is a variation of traditional supervised learning problems where data (referred to as bags) are composed of sub-elements (referred to as instances) and only bag labels are available.
no code implementations • NeurIPS 2016 • Kentaro Minami, Hitomi Arai, Issei Sato, Hiroshi Nakagawa
The exponential mechanism is a general method to construct a randomized estimator that satisfies $(\varepsilon, 0)$-differential privacy.
no code implementations • ICML 2018 • Weihua Hu, Gang Niu, Issei Sato, Masashi Sugiyama
Since the DRSL is explicitly formulated for a distribution shift scenario, we naturally expect it to give a robust classifier that can aggressively handle shifted distributions.
no code implementations • 4 Nov 2016 • Seiya Tokui, Issei Sato
Low-variance gradient estimation is crucial for learning directed graphical models parameterized by neural networks, where the reparameterization trick is widely used for those with continuous variables.
no code implementations • 10 Oct 2016 • Masatoshi Uehara, Issei Sato, Masahiro Suzuki, Kotaro Nakayama, Yutaka Matsuo
Generative adversarial networks (GANs) are successful deep generative models.
no code implementations • NeurIPS 2014 • Shinichi Nakajima, Issei Sato, Masashi Sugiyama, Kazuho Watanabe, Hiroko Kobayashi
Latent Dirichlet allocation (LDA) is a popular generative model of various objects such as texts and images, where an object is expressed as a mixture of latent topics.
no code implementations • 16 Sep 2014 • Katsuhiko Ishiguro, Issei Sato, Naonori Ueda
The Infinite Relational Model (IRM) is a probabilistic model for relational data clustering that partitions objects into clusters based on observed relationships.
no code implementations • 9 Aug 2014 • Issei Sato, Kenichi Kurihara, Shu Tanaka, Hiroshi Nakagawa, Seiji Miyashita
This paper presents studies on a deterministic annealing algorithm based on quantum annealing for variational Bayes (QAVB) inference, which can be seen as an extension of the simulated annealing for variational Bayes (SAVB) inference.
no code implementations • 19 May 2013 • Issei Sato, Shu Tanaka, Kenichi Kurihara, Seiji Miyashita, Hiroshi Nakagawa
We developed a new quantum annealing (QA) algorithm for Dirichlet process mixture (DPM) models based on the Chinese restaurant process (CRP).
no code implementations • NeurIPS 2010 • Issei Sato, Kenichi Kurihara, Hiroshi Nakagawa
We develop a deterministic single-pass algorithm for latent Dirichlet allocation (LDA) in order to process received documents one at a time and then discard them in an excess text stream.