This study considers multi-objective Bayesian optimization (MOBO) through the information gain of the Pareto-frontier. To calculate the information gain, a predictive distribution conditioned on the Pareto-frontier plays a key role, which is defined as a distribution truncated by the Pareto-frontier. However, it is usually impossible to obtain the entire Pareto-frontier in a continuous domain, and therefore, the complete truncation cannot be known. We consider an approximation of the truncated distribution by using a mixture distribution consisting of two possible approximate truncations obtainable from a subset of the Pareto-frontier, which we call over- and under-truncation. Since the optimal balance of the mixture is unknown beforehand, we propose optimizing the balancing coefficient through the variational lower bound maximization framework, by which the approximation error of the information gain can be minimized. Our empirical evaluation demonstrates the effectiveness of the proposed method particularly when the number of objective functions is large.

Gaussian process upper confidence bound (GP-UCB) is a theoretically established algorithm for Bayesian optimization (BO), where we assume the objective function ƒ follows a GP. One notable drawback of GP-UCB is that the theoretical confidence parameter β increases along with the iterations and is too large. To alleviate this drawback, this paper analyzes the randomized variant of GP-UCB called improved randomized GP-UCB (IRGP-UCB), which uses the confidence parameter generated from the shifted exponential distribution. We analyze the expected regret and conditional expected regret, where the expectation and the probability are taken respectively with ƒ and noise and with the randomness of the BO algorithm. In both regret analyses, IRGP-UCB achieves a sub-linear regret upper bound without increasing the confidence parameter if the input domain is finite. Furthermore, we show that randomization plays a key role in avoiding an increase in confidence parameter by showing that GP-UCB using a constant confidence parameter can incur linearly growing expected cumulative regret. Finally, we show numerical experiments using synthetic and benchmark functions and real-world emulators.

There are a lot of real-world black-box optimization problems that need to optimize multiple criteria simultaneously. However, in a multi-objective optimization (MOO) problem, identifying the whole Pareto front requires the prohibitive search cost, while in many practical scenarios, the decision maker (DM) only needs a specific solution among the set of the Pareto optimal solutions. We propose a Bayesian optimization (BO) approach to identifying the most preferred solution in the MOO with expensive objective functions, in which a Bayesian preference model of the DM is adaptively estimated by an interactive manner based on the two types of supervisions called the pairwise preference and improvement request. To explore the most preferred solution, we define an acquisition function in which the uncertainty both in the objective function and the DM preference is incorporated. Further, to minimize the interaction cost with the DM, we also propose an active learning strategy for the preference estimation. We empirically demonstrate the effectiveness of our proposed method through the benchmark function optimization and the hyper-parameter optimization problems for machine learning models.

Among various acquisition functions (AFs) in Bayesian optimization (BO), Gaussian process upper confidence bound (GP-UCB) and Thompson sampling (TS) are well-known options with established theoretical properties regarding Bayesian cumulative regret (BCR). Recently, it has been shown that a randomized variant of GP-UCB achieves a tighter BCR bound compared with GP-UCB, which we call the tighter BCR bound for brevity. Inspired by this study, this paper first shows that TS achieves the tighter BCR bound. On the other hand, GP-UCB and TS often practically suffer from manual hyperparameter tuning and over-exploration issues, respectively. Therefore, we analyze yet another AF called a probability of improvement from the maximum of a sample path (PIMS). We show that PIMS achieves the tighter BCR bound and avoids the hyperparameter tuning, unlike GP-UCB. Furthermore, we demonstrate a wide range of experiments, focusing on the effectiveness of PIMS that mitigates the practical issues of GP-UCB and TS.

The graph classification problem has been widely studied; however, achieving an interpretable model with high predictive performance remains a challenging issue. This paper proposes an interpretable classification algorithm for attributed graph data, called LAGRA (Learning Attributed GRAphlets). LAGRA learns importance weights for small attributed subgraphs, called attributed graphlets (AGs), while simultaneously optimizing their attribute vectors. This enables us to obtain a combination of subgraph structures and their attribute vectors that strongly contribute to discriminating different classes. A significant characteristics of LAGRA is that all the subgraph structures in the training dataset can be considered as a candidate structures of AGs. This approach can explore all the potentially important subgraphs exhaustively, but obviously, a naïve implementation can require a large amount of computations. To mitigate this issue, we propose an efficient pruning strategy by combining the proximal gradient descent and a graph mining tree search. Our pruning strategy can ensure that the quality of the solution is maintained compared to the result without pruning. We empirically demonstrate that LAGRA has superior or comparable prediction performance to the standard existing algorithms including graph neural networks, while using only a small number of AGs in an interpretable manner.

Bayesian optimization (BO) improves the efficiency of black-box optimization; however, the associated computational cost and power consumption remain dominant in the application of machine learning methods. This paper proposes a method of determining the stopping time in BO. The proposed criterion is based on the difference between the expectation of the minimum of a variant of the simple regrets before and after evaluating the objective function with a new parameter setting. Unlike existing stopping criteria, the proposed criterion is guaranteed to converge to the theoretically optimal stopping criterion for any choices of arbitrary acquisition functions and threshold values. Moreover, the threshold for the stopping criterion can be determined automatically and adaptively. We experimentally demonstrate that the proposed stopping criterion finds reasonable timing to stop a BO with a small number of evaluations of the objective function.

Gaussian process upper confidence bound (GP-UCB) is a theoretically promising approach for black-box optimization; however, the confidence parameter $\beta$ is considerably large in the theorem and chosen heuristically in practice. Then, randomized GP-UCB (RGP-UCB) uses a randomized confidence parameter, which follows the Gamma distribution, to mitigate the impact of manually specifying $\beta$. This study first generalizes the regret analysis of RGP-UCB to a wider class of distributions, including the Gamma distribution. Furthermore, we propose improved RGP-UCB (IRGP-UCB) based on a two-parameter exponential distribution, which achieves tighter Bayesian regret bounds. IRGP-UCB does not require an increase in the confidence parameter in terms of the number of iterations, which avoids over-exploration in the later iterations. Finally, we demonstrate the effectiveness of IRGP-UCB through extensive experiments.

We study preferential Bayesian optimization (BO) where reliable feedback is limited to pairwise comparison called duels. An important challenge in preferential BO, which uses the preferential Gaussian process (GP) model to represent flexible preference structure, is that the posterior distribution is a computationally intractable skew GP. The most widely used approach for preferential BO is Gaussian approximation, which ignores the skewness of the true posterior. Alternatively, Markov chain Monte Carlo (MCMC) based preferential BO is also proposed. In this work, we first verify the accuracy of Gaussian approximation, from which we reveal the critical problem that the predictive probability of duels can be inaccurate. This observation motivates us to improve the MCMC-based estimation for skew GP, for which we show the practical efficiency of Gibbs sampling and derive the low variance MC estimator. However, the computational time of MCMC can still be a bottleneck in practice. Towards building a more practical preferential BO, we develop a new method that achieves both high computational efficiency and low sample complexity, and then demonstrate its effectiveness through extensive numerical experiments.

In black-box function optimization, we need to consider not only controllable design variables but also uncontrollable stochastic environment variables. In such cases, it is necessary to solve the optimization problem by taking into account the uncertainty of the environmental variables. Chance-constrained (CC) problem, the problem of maximizing the expected value under a certain level of constraint satisfaction probability, is one of the practically important problems in the presence of environmental variables. In this study, we consider distributionally robust CC (DRCC) problem and propose a novel DRCC Bayesian optimization method for the case where the distribution of the environmental variables cannot be precisely specified. We show that the proposed method can find an arbitrary accurate solution with high probability in a finite number of trials, and confirm the usefulness of the proposed method through numerical experiments.

Max-value entropy search (MES) is one of the state-of-the-art approaches in Bayesian optimization (BO). In this paper, we propose a novel variant of MES for constrained problems, called Constrained MES via Information lower BOund (CMES-IBO), that is based on a Monte Carlo (MC) estimator of a lower bound of a mutual information (MI). Unlike existing studies, our MI is defined so that uncertainty with respect to feasibility can be incorporated. We derive a lower bound of the MI that guarantees non-negativity, while a constrained counterpart of conventional MES can be negative. We further provide theoretical analysis that assures the low-variability of our estimator which has never been investigated for any existing information-theoretic BO. Moreover, using the conditional MI, we extend CMES-IBO to the parallel setting while maintaining the desirable properties. We demonstrate the effectiveness of CMES-IBO by several benchmark functions and real-world problems.

Complex processes in science and engineering are often formulated as multistage decision-making problems. In this letter, we consider a cascade process, a type of multistage decision-making process. This is a multistage process in which the output of one stage is used as an input for the subsequent stage. When the cost of each stage is expensive, it is difficult to search for the optimal controllable parameters for each stage exhaustively. To address this problem, we formulate the optimization of the cascade process as an extension of the Bayesian optimization framework and propose two types of acquisition functions based on credible intervals and expected improvement. We investigate the theoretical properties of the proposed acquisition functions and demonstrate their effectiveness through numerical experiments. In addition, we consider suspension setting, an extension in which we are allowed to suspend the cascade process at the middle of the multistage decision-making process that often arises in practical problems. We apply the proposed method in a test problem involving a solar cell simulator, the motivation for this study.

Bayesian optimization (BO) is a popular method for expensive black-box optimization problems; however, querying the objective function at every iteration can be a bottleneck that hinders efficient search capabilities. In this regard, multifidelity Bayesian optimization (MFBO) aims to accelerate BO by incorporating lower-fidelity observations available with a lower sampling cost. In our previous work, we proposed an information-theoretic approach to MFBO, referred to as multifidelity max-value entropy search (MF-MES), which inherits practical effectiveness and computational simplicity of the well-known max-value entropy search (MES) for the single-fidelity BO. However, the applicability of MF-MES is still limited to the case that a single observation is sequentially obtained. In this letter, we generalize MF-MES so that information gain can be evaluated even when multiple observations are simultaneously obtained. This generalization enables MF-MES to address two practical problem settings: synchronous parallelization and trace-aware querying. We show that the acquisition functions for these extensions inherit the simplicity of MF-MES without introducing additional assumptions. We also provide computational techniques for entropy evaluation and posterior sampling in the acquisition functions, which can be commonly used for all variants of MF-MES. The effectiveness of MF-MES is demonstrated using benchmark functions and real-world applications such as materials science data and hyperparameter tuning of machine-learning algorithms.

We propose a novelstatistical approachto evaluate thestatistical significance(reliability) of the results from discriminative sub-trajectory mining, which we callStatistically Discriminative Sub-trajectory Mining (Stat-DSM). Given two groups of trajectories, the goal of Stat-DSM is to extract moving patterns in the form of sub-trajectories that occur statistically significantly more often in one group than in the other. An advantage of the proposed method is that the statistical significance of the extracted sub-trajectories are properly controlled in the sense that the probability of finding a falsely discriminative sub-trajectory is smaller than a specified significance threshold$\alpha$(e.g., 0.05), which is crucial when the method is used in scientific or social science studies under noisy environments. Finding such statistically discriminative sub-trajectories from a massive trajectory dataset is both computationally and statistically challenging. In the Stat-DSM method, we address these difficulties by introducing a tree representation of sub-trajectories, and applying an efficient permutation-based statistical inference method to the tree. To the best of our knowledge, Stat-DSM is the first method that provides a statistical approach to quantify the reliability of discriminative sub-trajectory mining results. We illustrate the effectiveness and scalability of the Stat-DSM method by applying it to a real-world dataset containing 1,000,000 trajectories.

This paper studies an entropy-based multi-objective Bayesian optimization (MBO). Existing entropy-based MBO methods need complicated approximations to evaluate entropy or employ over-simplification that ignores trade-off among objectives. We propose a novel entropy-based MBO called Pareto-frontier entropy search (PFES), which is based on the information gain of Pareto-frontier. We show that our entropy evaluation can be reduced to a closed form whose computation is quite simple while capturing the trade-off relation in Pareto-frontier. We further propose an extension for the “decoupled” setting, in which each objective function can be observed separately, and show that the PFES-based approach derives a natural extension of the original acquisition function which can also be evaluated simply. Our numerical experiments show effectiveness of PFES through several benchmark datasets, and real-word datasets from materials science.

In a standard setting of Bayesian optimization (BO), the objective function evaluation is assumed to be highly expensive. Multi-fidelity Bayesian optimization (MFBO) accelerates BO by incorporating lower fidelity observations available with a lower sampling cost. We propose a novel information-theoretic approach to MFBO, called multi-fidelity max-value entropy search (MF-MES), that enables us to obtain a more reliable evaluation of the information gain compared with existing information-based methods for MFBO. Further, we also propose a parallelization of MF-MES mainly for the asynchronous setting because queries typically occur asynchronously in MFBO due to a variety of sampling costs. We show that most of computations in our acquisition functions can be derived analytically, except for at most only two dimensional numerical integration that can be performed efficiently by simple approximations. We demonstrate effectiveness of our approach by using benchmark datasets and a real-world application to materials science data.

In this letter, we study an active learning problem for maximizing an unknown linear function with high-dimensional binary features. This problem is notoriously complex but arises in many important contexts. When the sampling budget, that is, the number of possible function evaluations, is smaller than the number of dimensions, it tends to be impossible to identify all of the optimal binary features. Therefore, in practice, only a small number of such features are considered, with the majority kept fixed at certain default values, which we call the working set heuristic. The main contribution of this letter is to formally study the working set heuristic and present a suite of theoretically robust algorithms for more efficient use of the sampling budget. Technically, we introduce a novel method for estimating the confidence regions of model parameters that is tailored to active learning with high-dimensional binary features. We provide a rigorous theoretical analysis of these algorithms and prove that a commonly used working set heuristic can identify optimal binary features with favorable sample complexity. We explore the performance of the proposed approach through numerical simulations and an application to a functional protein design problem.

Testing under what conditions a product satisfies the desired properties is a fundamental problem in manufacturing industry. If the condition and the property are respectively regarded as the input and the output of a black-box function, this task can be interpreted as the problem called level set estimation (LSE): the problem of identifying input regions such that the function value is above (or below) a threshold. Although various methods for LSE problems have been developed, many issues remain to be solved for their practical use. As one of such issues, we consider the case where the input conditions cannot be controlled precisely-LSE problems under input uncertainty. We introduce a basic framework for handling input uncertainty in LSE problems and then propose efficient methods with proper theoretical guarantees. The proposed methods and theories can be generally applied to a variety of challenges related to LSE under input uncertainty such as cost-dependent input uncertainties and unknown input uncertainties. We apply the proposed methods to artificial and real data to demonstrate their applicability and effectiveness.

We propose a novel statistical approach to evaluate the statistical significance (reliability) of the findings in the discriminative sub-trajectory mining problem, called Statistically Discriminative Sub-trajectory Mining (Stat-DSM). Given two groups of trajectories, the goal is to extract moving patterns in the form of sub-trajectories that occur statistically significantly more often in one group than in the other. An advantage of the Stat-DSM method is that the statistical significance of the extracted sub-trajectories are properly controlled in the sense that the probability of finding a false discriminative sub-trajectory is smaller than a specified significance threshold a (e.g., 0.05). We conduct experiments on real-world datasets to demonstrate the effectiveness of the Stat-DSM method.

Graph is a standard approach to modeling structured data. Although many machine learning methods depend on the metric of the input objects, defining an appropriate distance function on graph is still a controversial issue. We propose a novel supervised metric learning method for a subgraph-based distance, called interpretable graph metric learning (IGML). IGML optimizes the distance function in such a way that a small number of important subgraphs can be adaptively selected. This optimization is computationally intractable with naive application of existing optimization algorithms. We construct a graph mining based efficient algorithm to deal with this computational difficulty. Important advantages of our method are 1) guarantee of the optimality from the convex formulation, and 2) high interpretability of results. To our knowledge, none of the existing studies provide an interpretable subgraph-based metric in a supervised manner. In our experiments, we empirically verify superior or comparable prediction performance of IGML to other existing graph classification methods which do not have clear interpretability. Further, we demonstrate usefulness of IGML through some illustrative examples of extracted subgraphs and an example of data analysis on the learned metric space.

Distance metric learning has been widely used to obtain the optimal distance function based on the given training data. We focus on a triplet-based loss function, which imposes a penalty such that a pair of instances in the same class is closer than a pair in different classes. However, the number of possible triplets can be quite large even for a small data set, and this considerably increases the computational cost for metric optimization. In this letter, we propose safe triplet screening that identifies triplets that can be safely removed from the optimization problem without losing the optimality. In comparison with existing safe screening studies, triplet screening is particularly significant because of the huge number of possible triplets and the semidefinite constraint in the optimization problem. We demonstrate and verify the effectiveness of our screening rules by using several benchmark data sets.

Graph is a common way to represent relationships among a set of objects in a variety of application areas of machine learning. We consider the case that the input data is not only a graph but also numerical features in which one of the given features corresponds to a node in the graph. Then, the primary importance is often in understanding interactions on the graph nodes which effect on covariance structure of the numerical features. We propose a Gaussian based analysis which is a combination of graph constrained covariance matrix estimation and factor analysis (FA). We show that this approach, called graph FA, has desirable interpretability. In particular, we prove the connection between graph FA and a graph node clustering based on a perspective of kernel method. This connection indicates that graph FA is effective not only on the conventional noise-reduction explanation of the observation by FA but also on identifying important subgraphs. The experiments on synthetic and real-world datasets demonstrate the effectiveness of the approach.

We study safe screening for metric learning. Distance metric learning can optimize a metric over a set of triplets, each one of which is defined by a pair of same class instances and an instance in a different class. However, the number of possible triplets is quite huge even for a small dataset. Our safe triplet screening identifies triplets which can be safely removed from the optimization problem without losing the optimality. Compared with existing safe screening studies, triplet screening is particularly significant because of (1) the huge number of possible triplets, and (2) the semi-definite constraint in the optimization. We derive several variants of screening rules, and analyze their relationships. Numerical experiments on benchmark datasets demonstrate the effectiveness of safe triplet screening.

Triple-negative (TN) breast cancer (BC) patients have limited treatment options and poor prognosis even after extant treatments and standard chemotherapeutic regimens. Linking TN patients to clinically known phenotypes with appropriate treatments is vital. Location-specific sequence variants are expected to be useful for this purpose by identifying subgroups within a disease population. Single gene mutational signatures have been widely reported, with related phenotypes in literature. We thoroughly survey currently available mutations (and mutated genes), linked to BC phenotypes, to demonstrate their limited performance as sole predictors/biomarkers to assign phenotypes to patients. We then explore mutational combinations, as a pilot study, using The Cancer Genome Atlas Research Network mutational data of BC and three machine learning methods: association rules (limitless arity multiple procedure), decision tree and hierarchical disjoint clustering. The study results in a patient classification scheme through combinatorial mutations in Phosphatidylinositol-4,5-Bisphosphate 3-Kinase and tumor protein 53, being consistent with all three methods, implying its validity from a diverse viewpoint. However, it would warrant further research to select multi-gene signatures to identify phenotypes specifically and be clinically used routinely.

The problem of learning a sparse model is conceptually interpreted as the process of identifying active features/samples and then optimizing the model over them. Recently introduced safe screening allows us to identify a part of non-active features/samples. So far, safe screening has been individually studied either for feature screening or for sample screening. In this paper, we introduce a new approach for safely screening features and samples simultaneously by alternatively iterating feature and sample screening steps. A significant advantage of considering them simultaneously rather than individually is that they have a synergy effect in the sense that the results of the previous safe feature screening can be exploited for improving the next safe sample screening performances, and vice-versa. We first theoretically investigate the synergy effect, and then illustrate the practical advantage through intensive numerical experiments for problems with large numbers of features and samples.

In this paper we study predictive pattern mining problems where the goal is to construct a predictive model based on a subset of predictive patterns in the database. Our main contribution is to introduce a novel method called safe pattern pruning (SPP) for a class of predictive pattern mining problems. The SPP method allows us to efficiently find a superset of all the predictive patterns in the database that are needed for the optimal predictive model. The advantage of the SPP method over existing boosting-type method is that the former can find the superset by a single search over the database, while the latter requires multiple searches. The SPP method is inspired by recent development of safe feature screening. In order to extend the idea of safe feature screening into predictive pattern mining, we derive a novel pruning rule called safe pattern pruning (SPP) rule that can be used for searching over the tree defined among patterns in the database. The SPP rule has a property that, if a node corresponding to a pattern in the database is pruned out by the SPP rule, then it is guaranteed that all the patterns corresponding to its descendant nodes are never needed for the optimal predictive model. We apply the SPP method to graph mining and item-set mining problems, and demonstrate its computational advantage.

Careful tuning of a regularization parameter is indispensable in many machine learning tasks because it has a significant impact on generalization performances.Nevertheless, current practice of regularization parameter tuning is more of an art than a science, e.g., it is hard to tell how many grid-points would be needed in cross-validation (CV) for obtaining a solution with sufficiently small CV error.In this paper we propose a novel framework for computing a lower bound of the CV errors as a function of the regularization parameter, which we call regularization path of CV error lower bounds.The proposed framework can be used for providing a theoretical approximation guarantee on a set of solutions in the sense that how far the CV error of the current best solution could be away from best possible CV error in the entire range of the regularization parameters.We demonstrate through numerical experiments that a theoretically guaranteed a choice of regularization parameter in the above sense is possible with reasonable computational costs.

Label propagation is one of the state-of-the-art methods for semi-supervised learning, which estimates labels by propagating label information through a graph. Label propagation assumes that data points (nodes) connected in a graph should have similar labels. Consequently, the label estimation heavily depends on edge weights in a graph which represent similarity of each node pair. We propose a method for a graph to capture the manifold structure of input features using edge weights parameterized by a similarity function. In this approach, edge weights represent both similarity and local reconstruction weight simultaneously, both being reasonable for label propagation. For further justification, we provide analytical considerations including an interpretation as a cross-validation of a propagation model in the feature space, and an error analysis based on a low dimensional manifold model. Experimental results demonstrated the effectiveness of our approach both in synthetic and real datasets.

Graph-based approaches have been most successful in semisupervised learning. In this paper, we focus on label propagation in graph-based semisupervised learning. One essential point of label propagation is that the performance is heavily affected by incorporating underlying manifold of given data into the input graph. The other more important point is that in many recent real-world applications, the same instances are represented by multiple heterogeneous data sources. A key challenge under this setting is to integrate different data representations automatically to achieve better predictive performance. In this paper, we address the issue of obtaining the optimal linear combination of multiple different graphs under the label propagation setting. For this problem, we propose a new formulation with the sparsity (in coefficients of graph combination) property which cannot be rightly achieved by any other existing methods. This unique feature provides two important advantages: 1) the improvement of prediction performance by eliminating irrelevant or noisy graphs and 2) the interpretability of results, i.e., easily identifying informative graphs on classification. We propose efficient optimization algorithms for the proposed approach, by which clear interpretations of the mechanism for sparsity is provided. Through various synthetic and two real-world data sets, we empirically demonstrate the advantages of our proposed approach not only in prediction performance but also in graph selection ability.

Regularization path algorithms have been proposed to deal with model selection problem in several machine learning approaches. These algorithms allow computation of the entire path of solutions for every value of regularization parameter using the fact that their solution paths have piecewise linear form. In this paper, we extend the applicability of regularization path algorithm to a class of learning machines that have quadratic loss and quadratic penalty term. This class contains several important learning machines such as squared hinge loss support vector machine (SVM) and modified Huber loss SVM. We first show that the solution paths of this class of learning machines have piecewise nonlinear form, and piecewise segments between two breakpoints are characterized by a class of rational functions. Then we develop an algorithm that can efficiently follow the piecewise nonlinear path by solving these rational equations. To solve these rational equations, we use rational approximation technique with quadratic convergence rate, and thus, our algorithm can follow the nonlinear path much more precisely than existing approaches such as predictor-corrector type nonlinear-path approximation. We show the algorithm performance on some artificial and real data sets.

Regularization path algorithms have been proposed to deal with model selection problem in several machine learning approaches. These algorithms allow to compute the entire path of solutions for every value of regularization parameter using the fact that their solution paths have piecewise linear form. In this paper, we propose nonlinear regularization path for the Support Vector Machine (SVM) with a modified Huber loss. We first show that the solution path of the modified Huber loss SVM is represented as piecewise nonlinear function. Since the solutions between two breakpoints are characterized by a rational function, the breakpoint itself can be identified solving the rational equations. Then we develop an efficient iterative algorithm to solve these rational equations with quadratic convergence rate. Note that our algorithm is NOT a predictor-corrector type method that can only follow nonlinear regularization path with rough approximation. We show the algorithm performance on some artificial and real data sets

We propose a multiple incremental decremental algorithm of support vector machines (SVM). In online learning, we need to update the trained model when some new observations arrive and/or some observations become obsolete. If we want to add or remove single data point, conventional single incremental decremental algorithm can be used to update the model efficiently. However, to add and/or remove multiple data points, the computational cost of current update algorithm becomes inhibitive because we need to repeatedly apply it for each data point. In this paper, we develop an extension of incremental decremental algorithm which efficiently works for simultaneous update of multiple data points. Some analyses and experimental results show that the proposed algorithm can substantially reduce the computational cost. Our approach is especially useful for online SVM learning in which we need to remove old data points and add new data points in a short amount of time.

We propose a multiple incremental decremental algorithm of Support Vector Machine (SVM). Conventional single cremental decremental SVM can update the trained model efficiently when single data point is added to or removed from the training set. When we add and/or remove multiple data points, this algorithm is time-consuming because we need to repeatedly apply it to each data point. The roposed algorithm is computationally more efficient when multiple data points are added and/or removed simultaneously. The single incremental decremental algorithm is built on an optimization technique called parametric programming. We extend the idea and introduce multi-parametric programming for developing the proposed algorithm. Experimental results on synthetic and real data sets indicate that the proposed algorithm can significantly reduce the computational cost of multiple incremental decremental operation. Our approach is especially useful for online SVM learning in which we need to remove old data points and add new data points in a short amount of time.

Kernel ridge regression (KRR) is a nonlinear extension of the ridge regression. The performance of the KRR depends on its hyperparameters such as a penalty factor C, and RBF kernel parameter sigma. We employ a method called MCV-KRR which optimizes the KRR hyperparameters so that a cross-validation error is minimized. This method becomes equivalent to a predictive approach to Gaussian process. Since the cost of KRR training is O(N3) where N is a data size, to reduce this complexity, some sparse approximation of the KRR is recently studied. In this paper, we apply the minimum cross-validation (MCV) approach to such sparse approximation. Our experiments show the MCV with the sparse approximation of the KRR can achieve almost the same generalization performance as the MCV-KRR with much lower cost.

The performance of support vector regression (SVR) deeply depends on its hyperparameters such as an insensitive zone thickness, a penalty factor, and kernel parameters. A method called MCV-SVR was once proposed, which optimizes SVR hyperparameters so that cross-validation error is minimized. However, the computational cost of CV is usually high. In this paper we apply accurate online support vector regression (AOSVR) to the MCV-SVR cross-validation procedure. The AOSVR enables an efficient update of a trained SVR function when a sample is removed from training data. We show the AOSVR dramatically accelerates the MCV-SVR. Moreover, our experiments using real-world data showed our faster MCV-SVR has better generalization than other existing methods such as Bayesian SVR or practical setting.

The performance of support vector regression (SVR) deeply depends on its hyperparameters such as an insensitive zone thickness epsiv, a penalty factor C, and RBF kernel parameter sigma. A method called MCV-SVR was once proposed, which optimizes SVR hyperparameters so that a cross-validation error is minimized. However, as pointed out in this paper, the MCV-SVR (or its variants) has numerical instability in gradient calculation, which may cause bad influence on performance. Thus, this paper introduces a new method of computing the gradient of parameters with respect to hyperparameters. The revised optimizers incorporating the new method is shown to be free from the instability problem. Our experiments using three data sets showed that the revised optimizers considerably improved generalization performance of the MCV-SVR or its variant, and outperformed other methods such as multi-layer perceptrons or SVR with practical setting of hyperparameters.