Yongdai Kim

Dongyoon Yang, Insung Kong, Yongdai Kim

Adversarial training, which is to enhance robustness against adversarial attacks, has received much attention because it is easy to generate human-imperceptible perturbations of data to deceive a given deep neural network. In this paper, we propose a new adversarial training algorithm that is theoretically well motivated and empirically superior to other existing algorithms. A novel feature of the proposed algorithm is to apply more regularization to data vulnerable to adversarial attacks than other existing regularization algorithms do. Theoretically, we show that our algorithm can be understood as an algorithm of minimizing the regularized empirical risk motivated from a newly derived upper bound of the robust risk. Numerical experiments illustrate that our proposed algorithm improves the generalization (accuracy on examples) and robustness (accuracy on adversarial attacks) simultaneously to achieve the state-of-the-art performance.

Dongyoon Yang, Insung Kong, Yongdai Kim

Adversarial robustness is a research area that has recently received a lot of attention in the quest for trustworthy artificial intelligence. However, recent works on adversarial robustness have focused on supervised learning where it is assumed that labeled data is plentiful. In this paper, we investigate semi-supervised adversarial training where labeled data is scarce. We derive two upper bounds for the robust risk and propose a regularization term for unlabeled data motivated by these two upper bounds. Then, we develop a semi-supervised adversarial training algorithm that combines the proposed regularization term with knowledge distillation using a semi-supervised teacher (i.e., a teacher model trained using a semi-supervised learning algorithm). Our experiments show that our proposed algorithm achieves state-of-the-art performance with significant margins compared to existing algorithms. In particular, compared to supervised learning algorithms, performance of our proposed algorithm is not much worse even when the amount of labeled data is very small. For example, our algorithm with only 8\% labeled data is comparable to supervised adversarial training algorithms that use all labeled data, both in terms of standard and robust accuracies on CIFAR-10.

Dongha Kim, Jaesung Hwang, Jongjin Lee et al.

The unsupervised outlier detection (UOD) problem refers to a task to identify inliers given training data which contain outliers as well as inliers, without any labeled information about inliers and outliers. It has been widely recognized that using fully-trained likelihood-based deep generative models (DGMs) often results in poor performance in distinguishing inliers from outliers. In this study, we claim that the likelihood itself could serve as powerful evidence for identifying inliers in UOD tasks, provided that DGMs are carefully under-fitted. Our approach begins with a novel observation called the inlier-memorization (IM) effect-when training a deep generative model with data including outliers, the model initially memorizes inliers before outliers. Based on this finding, we develop a new method called the outlier detection via the IM effect (ODIM). Remarkably, the ODIM requires only a few updates, making it computationally efficient-at least tens of times faster than other deep-learning-based algorithms. Also, the ODIM filters out outliers excellently, regardless of the data type, including tabular, image, and text data. To validate the superiority and efficiency of our method, we provide extensive empirical analyses on close to 60 datasets.

Sara Kim, Kyusang Yu, Yongdai Kim

As they have a vital effect on social decision-making, AI algorithms not only should be accurate and but also should not pose unfairness against certain sensitive groups (e.g., non-white, women). Various specially designed AI algorithms to ensure trained AI models to be fair between sensitive groups have been developed. In this paper, we raise a new issue that between-group fair AI models could treat individuals in a same sensitive group unfairly. We introduce a new concept of fairness so-called within-group fairness which requires that AI models should be fair for those in a same sensitive group as well as those in different sensitive groups. We materialize the concept of within-group fairness by proposing corresponding mathematical definitions and developing learning algorithms to control within-group fairness and between-group fairness simultaneously. Numerical studies show that the proposed learning algorithms improve within-group fairness without sacrificing accuracy as well as between-group fairness.

Insung Kong, Dongyoon Yang, Jongjin Lee et al.

As data size and computing power increase, the architectures of deep neural networks (DNNs) have been getting more complex and huge, and thus there is a growing need to simplify such complex and huge DNNs. In this paper, we propose a novel sparse Bayesian neural network (BNN) which searches a good DNN with an appropriate complexity. We employ the masking variables at each node which can turn off some nodes according to the posterior distribution to yield a nodewise sparse DNN. We devise a prior distribution such that the posterior distribution has theoretical optimalities (i.e. minimax optimality and adaptiveness), and develop an efficient MCMC algorithm. By analyzing several benchmark datasets, we illustrate that the proposed BNN performs well compared to other existing methods in the sense that it discovers well condensed DNN architectures with similar prediction accuracy and uncertainty quantification compared to large DNNs.

Dongyoon Yang, Kunwoong Kim, Yongdai Kim

Semi-supervised learning (SSL) algorithm is a setup built upon a realistic assumption that access to a large amount of labeled data is tough. In this study, we present a generalized framework, named SCAR, standing for Selecting Clean samples with Adversarial Robustness, for improving the performance of recent SSL algorithms. By adversarially attacking pre-trained models with semi-supervision, our framework shows substantial advances in classifying images. We introduce how adversarial attacks successfully select high-confident unlabeled data to be labeled with current predictions. On CIFAR10, three recent SSL algorithms with SCAR result in significantly improved image classification.

Yongchan Choi, Seokhun Park, Chanmoo Park et al.

There are two things to be considered when we evaluate predictive models. One is prediction accuracy,and the other is interpretability. Over the recent decades, many prediction models of high performance, such as ensemble-based models and deep neural networks, have been developed. However, these models are often too complex, making it difficult to intuitively interpret their predictions. This complexity in interpretation limits their use in many real-world fields that require accountability, such as medicine, finance, and college admissions. In this study, we develop a novel method called Meta-ANOVA to provide an interpretable model for any given prediction model. The basic idea of Meta-ANOVA is to transform a given black-box prediction model to the functional ANOVA model. A novel technical contribution of Meta-ANOVA is a procedure of screening out unnecessary interaction before transforming a given black-box model to the functional ANOVA model. This screening procedure allows the inclusion of higher order interactions in the transformed functional ANOVA model without computational difficulties. We prove that the screening procedure is asymptotically consistent. Through various experiments with synthetic and real-world datasets, we empirically demonstrate the superiority of Meta-ANOVA

Seokhun Park, Choeun Kim, Kwanho Lee et al.

Activation functions play a central role in neural networks by shaping internal representations. Recently, learning binary activation representations has attracted significant attention due to their advantages in computational and memory efficiency, as well as interpretability. However, training neural networks with Heaviside activations remains challenging, as their non-differentiability obstructs standard gradient-based optimization. In this paper, we propose Heavy Tailed Activation Function (HTAF), a smooth approximation to the Heaviside function that enables stable training with gradient-based optimization. We construct HTAF as a sigmoid hyperbolic tangent composite function and theoretically show that it maintains a large gradient mass around zero inputs while exhibiting slower gradient decay in the tail regions. We show that Spiking Neural Networks, Binary Neural Networks and Deep Heaviside neural Networks can be trained stably using HTAF with gradient-based optimization. Finally, we introduce Implicit Concept Bottleneck Models (ICBMs), an interpretable image model that leverages HTAF to induce discrete feature representations. Extensive experiments across various architectures and image datasets demonstrate that ICBM enables stable discretization while achieving prediction performance comparable to or better than standard models.

Jinwon Park, Kunwoong Kim, Jihu Lee et al.

The goal of fair clustering is to find clusters such that the proportion of sensitive attributes (e.g., gender, race, etc.) in each cluster is similar to that of the entire dataset. Various fair clustering algorithms have been proposed that modify standard K-means clustering to satisfy a given fairness constraint. A critical limitation of several existing fair clustering algorithms is that the number of parameters to be learned is proportional to the sample size because the cluster assignment of each datum should be optimized simultaneously with the cluster center, and thus scaling up the algorithms is difficult. In this paper, we propose a new fair clustering algorithm based on a finite mixture model, called Fair Model-based Clustering (FMC). A main advantage of FMC is that the number of learnable parameters is independent of the sample size and thus can be scaled up easily. In particular, mini-batch learning is possible to obtain clusters that are approximately fair. Moreover, FMC can be applied to non-metric data (e.g., categorical data) as long as the likelihood is well-defined. Theoretical and empirical justifications for the superiority of the proposed algorithm are provided.

Insung Kong, Kunwoong Kim, Yongdai Kim

AI fairness, also known as algorithmic fairness, aims to ensure that algorithms operate without bias or discrimination towards any individual or group. Among various AI algorithms, the Fair Representation Learning (FRL) approach has gained significant interest in recent years. However, existing FRL algorithms have a limitation: they are primarily designed for categorical sensitive attributes and thus cannot be applied to continuous sensitive attributes, such as age or income. In this paper, we propose an FRL algorithm for continuous sensitive attributes. First, we introduce a measure called the Expectation of Integral Probability Metrics (EIPM) to assess the fairness level of representation space for continuous sensitive attributes. We demonstrate that if the distribution of the representation has a low EIPM value, then any prediction head constructed on the top of the representation become fair, regardless of the selection of the prediction head. Furthermore, EIPM possesses a distinguished advantage in that it can be accurately estimated using our proposed estimator with finite samples. Based on these properties, we propose a new FRL algorithm called Fair Representation using EIPM with MMD (FREM). Experimental evidences show that FREM outperforms other baseline methods.

Insung Kong, Yongdai Kim

Bayesian approaches for training deep neural networks (BNNs) have received significant interest and have been effectively utilized in a wide range of applications. There have been several studies on the properties of posterior concentrations of BNNs. However, most of these studies only demonstrate results in BNN models with sparse or heavy-tailed priors. Surprisingly, no theoretical results currently exist for BNNs using Gaussian priors, which are the most commonly used one. The lack of theory arises from the absence of approximation results of Deep Neural Networks (DNNs) that are non-sparse and have bounded parameters. In this paper, we present a new approximation theory for non-sparse DNNs with bounded parameters. Additionally, based on the approximation theory, we show that BNNs with non-sparse general priors can achieve near-minimax optimal posterior concentration rates to the true model.

Kunwoong Kim, Insung Kong, Jongjin Lee et al.

Group fairness requires that different protected groups, characterized by a given sensitive attribute, receive equal outcomes overall. Typically, the level of group fairness is measured by the statistical gap between predictions from different protected groups. In this study, we reveal an implicit property of existing group fairness measures, which provides an insight into how the group-fair models behave. Then, we develop a new group-fair constraint based on this implicit property to learn group-fair models. To do so, we first introduce a notable theoretical observation: every group-fair model has an implicitly corresponding transport map between the input spaces of each protected group. Based on this observation, we introduce a new group fairness measure termed Matched Demographic Parity (MDP), which quantifies the averaged gap between predictions of two individuals (from different protected groups) matched by a given transport map. Then, we prove that any transport map can be used in MDP to learn group-fair models, and develop a novel algorithm called Fairness Through Matching (FTM), which learns a group-fair model using MDP constraint with an user-specified transport map. We specifically propose two favorable types of transport maps for MDP, based on the optimal transport theory, and discuss their advantages. Experiments reveal that FTM successfully trains group-fair models with certain desirable properties by choosing the transport map accordingly.

Seokhun Park, Insung Kong, Yongchan Choi et al.

Interpretability for machine learning models is becoming more and more important as machine learning models become more complex. The functional ANOVA model, which decomposes a high-dimensional function into a sum of lower dimensional functions (commonly referred to as components), is one of the most popular tools for interpretable AI, and recently, various neural networks have been developed for estimating each component in the functional ANOVA model. However, such neural networks are highly unstable when estimating each component since the components themselves are not uniquely defined. That is, there are multiple functional ANOVA decompositions for a given function. In this paper, we propose a novel neural network which guarantees a unique functional ANOVA decomposition and thus is able to estimate each component stably and accurately. We call our proposed neural network ANOVA Tensor Product Neural Network (ANOVA-TPNN) since it is motivated by the tensor product basis expansion. Theoretically, we prove that ANOVA-TPNN can approximate any smooth function well. Empirically, we show that ANOVA-TPNN provide much more stable estimation of each component and thus much more stable interpretation when training data and initial values of the model parameters vary than existing neural networks do.

Kyungseon Lee, Kunwoong Kim, Jihu Lee et al.

Algorithmic fairness is a socially crucial topic in real-world applications of AI. Among many notions of fairness, subgroup fairness is widely studied when multiple sensitive attributes (e.g., gender, race, age) are present. However, as the number of sensitive attributes grows, the number of subgroups increases accordingly, creating heavy computational burdens and data sparsity problem (subgroups with too small sizes). In this paper, we develop a novel learning algorithm for subgroup fairness which resolves these issues by focusing on subgroups with sufficient sample sizes as well as marginal fairness (fairness for each sensitive attribute). To this end, we formalize a notion of subgroup-subset fairness and introduce a corresponding distributional fairness measure called the supremum Integral Probability Metric (supIPM). Building on this formulation, we propose the Doubly Regressing Adversarial learning for subgroup Fairness (DRAF) algorithm, which reduces a surrogate fairness gap for supIPM with much less computation than directly reducing supIPM. Theoretically, we prove that the proposed surrogate fairness gap is an upper bound of supIPM. Empirically, we show that the DRAF algorithm outperforms baseline methods in benchmark datasets, specifically when the number of sensitive attributes is large so that many subgroups are very small.

Sehyun Park, Jongjin Lee, Yunseop Shin et al.

Deep ensembles deliver state-of-the-art, reliable uncertainty quantification, but their heavy computational and memory requirements hinder their practical deployments to real applications such as on-device AI. Knowledge distillation compresses an ensemble into small student models, but existing techniques struggle to preserve uncertainty partly because reducing the size of DNNs typically results in variation reduction. To resolve this limitation, we introduce a new method of distribution distillation (i.e. compressing a teacher ensemble into a student distribution instead of a student ensemble) called Gaussian distillation, which estimates the distribution of a teacher ensemble through a special Gaussian process called the deep latent factor model (DLF) by treating each member of the teacher ensemble as a realization of a certain stochastic process. The mean and covariance functions in the DLF model are estimated stably by using the expectation-maximization (EM) algorithm. By using multiple benchmark datasets, we demonstrate that the proposed Gaussian distillation outperforms existing baselines. In addition, we illustrate that Gaussian distillation works well for fine-tuning of language models and distribution shift problems.

Seokhun Park, Choeun Kim, Jihu Lee et al.

With the increasing demand for interpretability in machine learning, functional ANOVA decomposition has gained renewed attention as a principled tool for breaking down high-dimensional function into low-dimensional components that reveal the contributions of different variable groups. Recently, Tensor Product Neural Network (TPNN) has been developed and applied as basis functions in the functional ANOVA model, referred to as ANOVA-TPNN. A disadvantage of ANOVA-TPNN, however, is that the components to be estimated must be specified in advance, which makes it difficult to incorporate higher-order TPNNs into the functional ANOVA model due to computational and memory constraints. In this work, we propose Bayesian-TPNN, a Bayesian inference procedure for the functional ANOVA model with TPNN basis functions, enabling the detection of higher-order components with reduced computational cost compared to ANOVA-TPNN. We develop an efficient MCMC algorithm and demonstrate that Bayesian-TPNN performs well by analyzing multiple benchmark datasets. Theoretically, we prove that the posterior of Bayesian-TPNN is consistent.

Seokhun Park, Insung Kong, Yongdai Kim

Bayesian Additive Regression Trees (BART) is a powerful statistical model that leverages the strengths of Bayesian inference and regression trees. It has received significant attention for capturing complex non-linear relationships and interactions among predictors. However, the accuracy of BART often comes at the cost of interpretability. To address this limitation, we propose ANOVA Bayesian Additive Regression Trees (ANOVA-BART), a novel extension of BART based on the functional ANOVA decomposition, which is used to decompose the variability of a function into different interactions, each representing the contribution of a different set of covariates or factors. Our proposed ANOVA-BART enhances interpretability, preserves and extends the theoretical guarantees of BART, and achieves superior predictive performance. Specifically, we establish that the posterior concentration rate of ANOVA-BART is nearly minimax optimal, and further provides the same convergence rates for each interaction that are not available for BART. Moreover, comprehensive experiments confirm that ANOVA-BART surpasses BART in both accuracy and uncertainty quantification, while also demonstrating its effectiveness in component selection. These results suggest that ANOVA-BART offers a compelling alternative to BART by balancing predictive accuracy, interpretability, and theoretical consistency.

Jihu Lee, Kunwoong Kim, Yongdai Kim

Fair clustering has become a socially significant task with the advancement of machine learning technologies and the growing demand for trustworthy AI. Group fairness ensures that the proportions of each sensitive group are similar in all clusters. Most existing group-fair clustering methods are based on the $K$-means clustering and thus require the distance between instances and the number of clusters to be given in advance. To resolve this limitation, we propose a fair Bayesian model-based clustering called Fair Bayesian Clustering (FBC). We develop a specially designed prior which puts its mass only on fair clusters, and implement an efficient MCMC algorithm. Advantages of FBC are that it can infer the number of clusters and can be applied to any data type as long as the likelihood is defined (e.g., categorical data). Experiments on real-world datasets show that FBC (i) reasonably infers the number of clusters, (ii) achieves a competitive utility-fairness trade-off compared to existing fair clustering methods, and (iii) performs well on categorical data.

Kunwoong Kim, Jihu Lee, Sangchul Park et al.

Algorithmic fairness in clustering aims to balance the proportions of instances assigned to each cluster with respect to a given sensitive attribute. While recently developed fair clustering algorithms optimize clustering objectives under specific fairness constraints, their inherent complexity or approximation often results in suboptimal clustering utility or numerical instability in practice. To resolve these limitations, we propose a new fair clustering algorithm based on a novel decomposition of the fair $K$-means clustering objective function. The proposed algorithm, called Fair Clustering via Alignment (FCA), operates by alternately (i) finding a joint probability distribution to align the data from different protected groups, and (ii) optimizing cluster centers in the aligned space. A key advantage of FCA is that it theoretically guarantees approximately optimal clustering utility for any given fairness level without complex constraints, thereby enabling high-utility fair clustering in practice. Experiments show that FCA outperforms existing methods by (i) attaining a superior trade-off between fairness level and clustering utility, and (ii) achieving near-perfect fairness without numerical instability.

Dongyoon Yang, Jihu Lee, Yongdai Kim

Robust domain adaptation against adversarial attacks is a critical research area that aims to develop models capable of maintaining consistent performance across diverse and challenging domains. In this paper, we derive a new generalization bound for robust risk on the target domain using a novel divergence measure specifically designed for robust domain adaptation. Building upon this, we propose a new algorithm named TAROT, which is designed to enhance both domain adaptability and robustness. Through extensive experiments, TAROT not only surpasses state-of-the-art methods in accuracy and robustness but also significantly enhances domain generalization and scalability by effectively learning domain-invariant features. In particular, TAROT achieves superior performance on the challenging DomainNet dataset, demonstrating its ability to learn domain-invariant representations that generalize well across different domains, including unseen ones. These results highlight the broader applicability of our approach in real-world domain adaptation scenarios.

Yuha Park, Kunwoong Kim, Insung Kong et al.

We propose a parametric integral probability metric (IPM) to measure the discrepancy between two probability measures. The proposed IPM leverages a specific parametric family of discriminators, such as single-node neural networks with ReLU activation, to effectively distinguish between distributions, making it applicable in high-dimensional settings. By optimizing over the parameters of the chosen discriminator class, the proposed IPM demonstrates that its estimators have good convergence rates and can serve as a surrogate for other IPMs that use smooth nonparametric discriminator classes. We present an efficient algorithm for practical computation, offering a simple implementation and requiring fewer hyperparameters. Furthermore, we explore its applications in various tasks, such as covariate balancing for causal inference and fair representation learning. Across such diverse applications, we demonstrate that the proposed IPM provides strong theoretical guarantees, and empirical experiments show that it achieves comparable or even superior performance to other methods.

Ilsang Ohn, Lizhen Lin, Yongdai Kim

In this paper, we propose a new Bayesian inference method for a high-dimensional sparse factor model that allows both the factor dimensionality and the sparse structure of the loading matrix to be inferred. The novelty is to introduce a certain dependence between the sparsity level and the factor dimensionality, which leads to adaptive posterior concentration while keeping computational tractability. We show that the posterior distribution asymptotically concentrates on the true factor dimensionality, and more importantly, this posterior consistency is adaptive to the sparsity level of the true loading matrix and the noise variance. We also prove that the proposed Bayesian model attains the optimal detection rate of the factor dimensionality in a more general situation than those found in the literature. Moreover, we obtain a near-optimal posterior concentration rate of the covariance matrix. Numerical studies are conducted and show the superiority of the proposed method compared with other competitors.

Insung Kong, Dongyoon Yang, Jongjin Lee et al.

Bayesian approaches for learning deep neural networks (BNN) have been received much attention and successfully applied to various applications. Particularly, BNNs have the merit of having better generalization ability as well as better uncertainty quantification. For the success of BNN, search an appropriate architecture of the neural networks is an important task, and various algorithms to find good sparse neural networks have been proposed. In this paper, we propose a new node-sparse BNN model which has good theoretical properties and is computationally feasible. We prove that the posterior concentration rate to the true model is near minimax optimal and adaptive to the smoothness of the true model. In particular the adaptiveness is the first of its kind for node-sparse BNNs. In addition, we develop a novel MCMC algorithm which makes the Bayesian inference of the node-sparse BNN model feasible in practice.

Insung Kong, Yuha Park, Joonhyuk Jung et al.

Weighting methods in causal inference have been widely used to achieve a desirable level of covariate balancing. However, the existing weighting methods have desirable theoretical properties only when a certain model, either the propensity score or outcome regression model, is correctly specified. In addition, the corresponding estimators do not behave well for finite samples due to large variance even when the model is correctly specified. In this paper, we consider to use the integral probability metric (IPM), which is a metric between two probability measures, for covariate balancing. Optimal weights are determined so that weighted empirical distributions for the treated and control groups have the smallest IPM value for a given set of discriminators. We prove that the corresponding estimator can be consistent without correctly specifying any model (neither the propensity score nor the outcome regression model). In addition, we empirically show that our proposed method outperforms existing weighting methods with large margins for finite samples.

Kunwoong Kim, Ilsang Ohn, Sara Kim et al.

As they have a vital effect on social decision makings, AI algorithms should be not only accurate and but also fair. Among various algorithms for fairness AI, learning a prediction model by minimizing the empirical risk (e.g., cross-entropy) subject to a given fairness constraint has received much attention. To avoid computational difficulty, however, a given fairness constraint is replaced by a surrogate fairness constraint as the 0-1 loss is replaced by a convex surrogate loss for classification problems. In this paper, we investigate the validity of existing surrogate fairness constraints and propose a new surrogate fairness constraint called SLIDE, which is computationally feasible and asymptotically valid in the sense that the learned model satisfies the fairness constraint asymptotically and achieves a fast convergence rate. Numerical experiments confirm that the SLIDE works well for various benchmark datasets.

Dongha Kim, Kunwoong Kim, Insung Kong et al.

As they have a vital effect on social decision-making, AI algorithms should be not only accurate but also fair. Among various algorithms for fairness AI, learning fair representation (LFR), whose goal is to find a fair representation with respect to sensitive variables such as gender and race, has received much attention. For LFR, the adversarial training scheme is popularly employed as is done in the generative adversarial network type algorithms. The choice of a discriminator, however, is done heuristically without justification. In this paper, we propose a new adversarial training scheme for LFR, where the integral probability metric (IPM) with a specific parametric family of discriminators is used. The most notable result of the proposed LFR algorithm is its theoretical guarantee about the fairness of the final prediction model, which has not been considered yet. That is, we derive theoretical relations between the fairness of representation and the fairness of the prediction model built on the top of the representation (i.e., using the representation as the input). Moreover, by numerical experiments, we show that our proposed LFR algorithm is computationally lighter and more stable, and the final prediction model is competitive or superior to other LFR algorithms using more complex discriminators.

Dongha Kim, Yongchan Choi, Kunwoong Kim et al.

In many classification problems, collecting massive clean-annotated data is not easy, and thus a lot of researches have been done to handle data with noisy labels. Most recent state-of-art solutions for noisy label problems are built on the small-loss strategy which exploits the memorization effect. While it is a powerful tool, the memorization effect has several drawbacks. The performances are sensitive to the choice of a training epoch required for utilizing the memorization effect. In addition, when the labels are heavily contaminated or imbalanced, the memorization effect may not occur in which case the methods based on the small-loss strategy fail to identify clean labeled data. We introduce a new method called INN(Integration with the Nearest Neighborhoods) to refine clean labeled data from training data with noisy labels. The proposed method is based on a new discovery that a prediction pattern at neighbor regions of clean labeled data is consistently different from that of noisy labeled data regardless of training epochs. The INN method requires more computation but is much stable and powerful than the small-loss strategy. By carrying out various experiments, we demonstrate that the INN method resolves the shortcomings in the memorization effect successfully and thus is helpful to construct more accurate deep prediction models with training data with noisy labels.

Minwoo Chae, Dongha Kim, Yongdai Kim et al.

We investigate statistical properties of a likelihood approach to nonparametric estimation of a singular distribution using deep generative models. More specifically, a deep generative model is used to model high-dimensional data that are assumed to concentrate around some low-dimensional structure. Estimating the distribution supported on this low-dimensional structure, such as a low-dimensional manifold, is challenging due to its singularity with respect to the Lebesgue measure in the ambient space. In the considered model, a usual likelihood approach can fail to estimate the target distribution consistently due to the singularity. We prove that a novel and effective solution exists by perturbing the data with an instance noise, which leads to consistent estimation of the underlying distribution with desirable convergence rates. We also characterize the class of distributions that can be efficiently estimated via deep generative models. This class is sufficiently general to contain various structured distributions such as product distributions, classically smooth distributions and distributions supported on a low-dimensional manifold. Our analysis provides some insights on how deep generative models can avoid the curse of dimensionality for nonparametric distribution estimation. We conduct a thorough simulation study and real data analysis to empirically demonstrate that the proposed data perturbation technique improves the estimation performance significantly.

Minjin Kim, Young-geun Kim, Dongha Kim et al.

The Mixup method (Zhang et al. 2018), which uses linearly interpolated data, has emerged as an effective data augmentation tool to improve generalization performance and the robustness to adversarial examples. The motivation is to curtail undesirable oscillations by its implicit model constraint to behave linearly at in-between observed data points and promote smoothness. In this work, we formally investigate this premise, propose a way to explicitly impose smoothness constraints, and extend it to incorporate with implicit model constraints. First, we derive a new function class composed of kernel-convoluted models (KCM) where the smoothness constraint is directly imposed by locally averaging the original functions with a kernel function. Second, we propose to incorporate the Mixup method into KCM to expand the domains of smoothness. In both cases of KCM and the KCM adapted with the Mixup, we provide risk analysis, respectively, under some conditions for kernels. We show that the upper bound of the excess risk is not slower than that of the original function class. The upper bound of the KCM with the Mixup remains dominated by that of the KCM if the perturbation of the Mixup vanishes faster than \(O(n^{-1/2})\) where \(n\) is a sample size. Using CIFAR-10 and CIFAR-100 datasets, our experiments demonstrate that the KCM with the Mixup outperforms the Mixup method in terms of generalization and robustness to adversarial examples.

Ilsang Ohn, Yongdai Kim

Recent theoretical studies proved that deep neural network (DNN) estimators obtained by minimizing empirical risk with a certain sparsity constraint can attain optimal convergence rates for regression and classification problems. However, the sparsity constraint requires to know certain properties of the true model, which are not available in practice. Moreover, computation is difficult due to the discrete nature of the sparsity constraint. In this paper, we propose a novel penalized estimation method for sparse DNNs, which resolves the aforementioned problems existing in the sparsity constraint. We establish an oracle inequality for the excess risk of the proposed sparse-penalized DNN estimator and derive convergence rates for several learning tasks. In particular, we prove that the sparse-penalized estimator can adaptively attain minimax convergence rates for various nonparametric regression problems. For computation, we develop an efficient gradient-based optimization algorithm that guarantees the monotonic reduction of the objective function.

Dongha Kim, Yongchan Choi, Yongdai Kim

In semi-supervised learning, virtual adversarial training (VAT) approach is one of the most attractive method due to its intuitional simplicity and powerful performances. VAT finds a classifier which is robust to data perturbation toward the adversarial direction. In this study, we provide a fundamental explanation why VAT works well in semi-supervised learning case and propose new techniques which are simple but powerful to improve the VAT method. Especially we employ the idea of Bad GAN approach, which utilizes bad samples distributed on complement of the support of the input data, without any additional deep generative architectures. We generate bad samples of high-quality by use of the adversarial training used in VAT and also give theoretical explanations why the adversarial training is good at both generating bad samples. An advantage of our proposed method is to achieve the competitive performances compared with other recent studies with much fewer computations. We demonstrate advantages our method by various experiments with well known benchmark image datasets.

Ilsang Ohn, Yongdai Kim

There has been a growing interest in expressivity of deep neural networks. However, most of the existing work about this topic focuses only on the specific activation function such as ReLU or sigmoid. In this paper, we investigate the approximation ability of deep neural networks with a broad class of activation functions. This class of activation functions includes most of frequently used activation functions. We derive the required depth, width and sparsity of a deep neural network to approximate any Hölder smooth function upto a given approximation error for the large class of activation functions. Based on our approximation error analysis, we derive the minimax optimality of the deep neural network estimators with the general activation functions in both regression and classification problems.

Jong-June Jeon, Yongdai Kim, Sungho Won et al.

Compositional data have two unique characteristics compared to typical multivariate data: the observed values are nonnegative and their summand is exactly one. To reflect these characteristics, a specific regularized regression model with linear constraints is commonly used. However, linear constraints incur additional computational time, which becomes severe in high-dimensional cases. As such, we propose an efficient solution path algorithm for a $l_1$ regularized regression with compositional data. The algorithm is then extended to a classification model with compositional predictors. We also compare its computational speed with that of previously developed algorithms and apply the proposed algorithm to analyze human gut microbiome data.

Yongdai Kim, Ilsang Ohn, Dongha Kim

We derive the fast convergence rates of a deep neural network (DNN) classifier with the rectified linear unit (ReLU) activation function learned using the hinge loss. We consider three cases for a true model: (1) a smooth decision boundary, (2) smooth conditional class probability, and (3) the margin condition (i.e., the probability of inputs near the decision boundary is small). We show that the DNN classifier learned using the hinge loss achieves fast rate convergences for all three cases provided that the architecture (i.e., the number of layers, number of nodes and sparsity). is carefully selected. An important implication is that DNN architectures are very flexible for use in various cases without much modification. In addition, we consider a DNN classifier learned by minimizing the cross-entropy, and show that the DNN classifier achieves a fast convergence rate under the condition that the conditional class probabilities of most data are sufficiently close to either 1 or zero. This assumption is not unusual for image recognition because human beings are extremely good at recognizing most images. To confirm our theoretical explanation, we present the results of a small numerical study conducted to compare the hinge loss and cross-entropy.

Yongdai Kim, Dongha Kim

We provide a theoretical explanation of the role of the number of nodes at each layer in deep neural networks. We prove that the largest variation of a deep neural network with ReLU activation function arises when the layer with the fewest nodes changes its activation pattern. An important implication is that deep neural network is a useful tool to generate functions most of whose variations are concentrated on a smaller area of the input space near the boundaries corresponding to the layer with the fewest nodes. In turn, this property makes the function more invariant to input transformation. That is, our theoretical result gives a clue about how to design the architecture of a deep neural network to increase complexity and transformation invariancy simultaneously.