arXiv subjects

Ken Kobayashi

Publications and source records attributed to Ken Kobayashi.

At least 19 recordsLinked to original sources

arXiv2026

Distributionally robust optimization for recommendation selection

Recommender systems play an essential role in online services by providing personalized item lists to support users' decision-making processes. While collaborative filtering methods can achieve high accuracy, it is crucial to consider not only accuracy but also the diversity of recommended items to improve user satisfaction. Although financial portfolio theory has been applied to balance these factors, existing models are often sensitive to estimation errors in rating statistics. To overcome these challenges, we establish a computational framework of distributionally robust optimization (DRO) for recommendation selection. We first formulate a cardinality-constrained DRO model based on moment-based ambiguity sets to select a specified number of items for each user. We then design a penalty alternating direction method (PADM) to efficiently compute high-quality solutions and prove its convergence properties. Computational experiments using three publicly available rating datasets demonstrate that our DRO model generates more diverse recommendations than existing models while maintaining the same level of accuracy. Additionally, our solution method computes these recommendations for each user in just a few seconds, proving its practical effectiveness. This study establishes a DRO framework that has the potential to enhance the recommendation quality of various collaborative filtering methods.

math.OC↗

arXiv2026

Kd-tree Based Wasserstein Distance Approximation for High-Dimensional Data

The Wasserstein distance is a discrepancy measure between probability distributions, defined by an optimal transport problem. It has been used for various tasks such as retrieving similar items in high-dimensional images or text data. In retrieval applications, however, the Wasserstein distance is calculated repeatedly, and its cubic time complexity with respect to input size renders it unsuitable for large-scale datasets. Recently, tree-based approximation methods have been proposed to address this bottleneck. For example, the Flowtree algorithm computes transport on a quadtree and evaluates cost using the ground metric, and clustering-tree approaches have been reported to achieve high accuracy. However, these existing trees often incur significant construction time for preprocessing, and crucially, standard quadtrees cannot grow deep enough in high-dimensional spaces, resulting in poor approximation accuracy. In this paper, we propose kd-Flowtree, a kd-tree-based Wasserstein distance approximation method that uses a kd-tree for data embedding. Since kd-trees can grow sufficiently deep and adaptively even in high-dimensional cases, kd-Flowtree is capable of maintaining good approximation accuracy for such cases. In addition, kd-trees can be constructed quickly than quadtrees, which contributes to reducing the computation time required for nearest neighbor search, including preprocessing. We provide a probabilistic upper bound on the nearest-neighbor search accuracy of kd-Flowtree, and show that this bound is independent of the dataset size. In the numerical experiments, we demonstrated that kd-Flowtree outperformed the existing Wasserstein distance approximation methods for retrieval tasks with real-world data.

cs.DS↗

arXiv2025

Content-Aware Ad Banner Layout Generation with Two-Stage Chain-of-Thought in Vision Language Models

In this paper, we propose a method for generating layouts for image-based advertisements by leveraging a Vision-Language Model (VLM). Conventional advertisement layout techniques have predominantly relied on saliency mapping to detect salient regions within a background image, but such approaches often fail to fully account for the image's detailed composition and semantic content. To overcome this limitation, our method harnesses a VLM to recognize the products and other elements depicted in the background and to inform the placement of text and logos. The proposed layout-generation pipeline consists of two steps. In the first step, the VLM analyzes the image to identify object types and their spatial relationships, then produces a text-based "placement plan" based on this analysis. In the second step, that plan is rendered into the final layout by generating HTML-format code. We validated the effectiveness of our approach through evaluation experiments, conducting both quantitative and qualitative comparisons against existing methods. The results demonstrate that by explicitly considering the background image's content, our method produces noticeably higher-quality advertisement layouts.

cs.CV↗

arXiv2025

A sudden fine-scale bright kernel captured by Hi-C Flare in 11 MK emission during an M1.6-class solar flare's post-maximum phase

On April 17, 2024, the third successful Hi-C sounding rocket flight, Hi-C Flare, recorded coronal images in Fe XXI 129 A emission from 11 MK plasma during the post-maximum phase of an M1.6-class solar flare, achieving unprecedented spatial (~300 km) and temporal (1.3 s) resolutions. The flare started at 21:55 UT, peaked at 22:08 UT, and lasted ~40 minutes. Hi-C observed for over five minutes (22:15:45 to 22:21:25), starting roughly eight minutes after flare maximum. A sudden compact bright burst - 875 +/- 25 km wide, lasting 90 +/- 1.3 s, exhibiting a proper motion of ~50 km/s, and splitting into two toward the end - occurs near the foot of some post-flare loops. Its size and brightness are reminiscent of flare-ribbon kernels during a flare's rapid rise phase, kernels marking sites of sudden heating and hot plasma upflow, making its occurrence during the late phase surprising. Such isolated brightenings in a flare's post-maximum phase are rare, and have not been previously reported. The kernel was detected in all SDO/AIA channels. Its 1600 A light curve peaked ~50 s earlier than its 131 A light curve, similar to that of flare-ribbon kernels, albeit with a smaller delay of ~25 s, during the impulsive phase of the flare. In SDO/HMI magnetograms, the kernel sits in unipolar positive magnetic flux near an embedded clump of negative flux. Although localized magnetic reconnection within the kernel (a microflare) cannot be ruled out for its cause, the observations favor the localized brightening being an isolated, exceptionally late flare-ribbon kernel, resulting from an exceptionally late burst of the flare's coronal reconnection.

astro-ph.SR↗

arXiv2025

Hierarchical Time Series Forecasting with Robust Reconciliation

This paper focuses on forecasting hierarchical time-series data, where each higher-level observation equals the sum of its corresponding lower-level time series. In such contexts, the forecast values should be coherent, meaning that the forecast value of each parent series exactly matches the sum of the forecast values of its child series. Existing hierarchical forecasting methods typically generate base forecasts independently for each series and then apply a reconciliation procedure to adjust them so that the resulting forecast values are coherent across the hierarchy. These methods generally derive an optimal reconciliation, using a covariance matrix of the forecast error. In practice, however, the true covariance matrix is unknown and has to be estimated from finite samples in advance. This gap between the true and estimated covariance matrix may degrade forecast performance. To address this issue, we propose a robust optimization framework for hierarchical reconciliation that accounts for uncertainty in the estimated covariance matrix. We first introduce an uncertainty set for the estimated covariance matrix and formulate a reconciliation problem that minimizes the worst-case average of weighted squared residuals over this uncertainty set. We show that our problem can be cast as a semidefinite optimization problem. Numerical experiments demonstrate that the proposed robust reconciliation method achieved better forecast performance than existing hierarchical forecasting methods, which indicates the effectiveness of integrating uncertainty into the reconciliation process.

cs.LG↗

arXiv2025

Determining the magnetic field of active region plages using the whole CLASP2/2.1 spectral window

The Chromospheric LAyer SpectroPolarimeter missions, CLASP2 and CLASP2.1, demonstrated that the near-UV spectral region between 279.30 and 280.68 nm is suitable for studying the magnetism of the solar chromosphere. In particular, the spectropolarimetric observations in the Mg II h and k resonant doublet, Mn I 279.91 and 280.19 nm resonant lines, and Fe II 279.79 and 280.66 nm lines acquired by these suborbital space experiments have been proven useful for inferring the magnetic field stratification in the solar chromosphere. However, several lines of the CLASP2/2.1 spectral region with significant circular polarization signals had remained unexplored. After identifying two Ni I (279.95 and 280.59 nm), one Mn II (280.62 nm), and one Fe I (280.53 nm) lines, here we apply the Weak Field Approximation (WFA) to the spectropolarimetric observations of active region plages by CLASP2 and CLASP2.1. By comparing the results with previous studies, we are able to estimate the formation heights of these CLASP2/2.1 additional spectral lines and to demonstrate their suitability to determine the magnetic field stratification from the photosphere to the upper chromosphere.

astro-ph.SR↗

arXiv2025

Determining the Magnetic Field in the Atmosphere of a Solar Active Region Observed by the CLASP2.1 Sounding Rocket Experiment

We determine magnetic fields from the photosphere to the upper chromosphere combining data from the Hinode satellite and the CLASP2.1 sounding rocket experiment. CLASP2.1 provided polarization profiles of the Mg~{|sc ii} $h$ and $k$ lines, as well as of the Mn~{|sc i} lines around 2800~{|AA}, across various magnetic structures in an active region, containing a plage, a pore, and the edges of a sunspot penumbra. By applying the Weak-Field Approximation (WFA) to the circular polarization profiles of these spectral lines, we obtain a longitudinal magnetic field map at three different heights in the chromosphere (lower, middle, and upper). This is complemented by data from Hinode (photospheric magnetic field), IRIS, and SDO (high-spatial-resolution observations of the chromosphere and corona). We quantify the height expansion of the plage magnetic fields and find that the magnetic fields expand significantly in the middle chromosphere, shaping the moss observed above in the transition region and corona. We identified an area with polarity reversal at the upper chromosphere around the edge of the pore, suggesting the presence of a magnetic discontinuity in the upper chromosphere. Transient and recurrent jet-like events are observed in this region, likely driven by magnetic reconnection. Around the penumbral edge, we find large-scale magnetic fields corresponding to the superpenumbral fibrils seen in the upper chromosphere. In the superpenumbral fibrils, we find Zeeman-induced linear polarization signals, suggesting the presence of a significantly inclined magnetic field, as strong as 1000~G in the upper chromosphere.

astro-ph.SR↗

arXiv2025

Interior-Point Vanishing Problem in Semidefinite Relaxations for Neural Network Verification

Semidefinite programming (SDP) relaxation has emerged as a promising approach for neural network verification, offering tighter bounds than other convex relaxation methods for deep neural networks (DNNs) with ReLU activations. However, we identify a critical limitation in the SDP relaxation when applied to deep networks: interior-point vanishing, which leads to the loss of strict feasibility -- a crucial condition for the numerical stability and optimality of SDP. Through rigorous theoretical and empirical analysis, we demonstrate that as the depth of DNNs increases, the strict feasibility is likely to be lost, creating a fundamental barrier to scaling SDP-based verification. To address the interior-point vanishing, we design and investigate five solutions to enhance the feasibility conditions of the verification problem. Our methods can successfully solve 88% of the problems that could not be solved by existing methods, accounting for 41% of the total. Our analysis also reveals that the valid constraints for the lower and upper bounds for each ReLU unit are traditionally inherited from prior work without solid reasons, but are actually not only unbeneficial but also even harmful to the problem's feasibility. This work provides valuable insights into the fundamental challenges of SDP-based DNN verification and offers practical solutions to improve its applicability to deeper neural networks, contributing to the development of more reliable and secure systems with DNNs.

cs.LG↗

arXiv2024

Balancing Immediate Revenue and Future Off-Policy Evaluation in Coupon Allocation

Coupon allocation drives customer purchases and boosts revenue. However, it presents a fundamental trade-off between exploiting the current optimal policy to maximize immediate revenue and exploring alternative policies to collect data for future policy improvement via off-policy evaluation (OPE). To balance this trade-off, we propose a novel approach that combines a model-based revenue maximization policy and a randomized exploration policy for data collection. Our framework enables flexible adjustment of the mixture ratio between these two policies to optimize the balance between short-term revenue and future policy improvement. We formulate the problem of determining the optimal mixture ratio as multi-objective optimization, enabling quantitative evaluation of this trade-off. We empirically verified the effectiveness of the proposed mixed policy using synthetic data. Our main contributions are: (1) Demonstrating a mixed policy combining deterministic and probabilistic policies, flexibly adjusting the data collection vs. revenue trade-off. (2) Formulating the optimal mixture ratio problem as multi-objective optimization, enabling quantitative evaluation of this trade-off.

cs.LG↗

arXiv2024

Mapping the longitudinal magnetic field in the atmosphere of an active region plage from the inversion of the near-ultraviolet CLASP2.1 spectropolarimetric data

We apply the HanleRT Tenerife Inversion Code to the spectro-polarimetric observations obtained by the Chromospheric LAyer SpectroPolarimeter. This suborbital space experiment measured the variation with wavelength of the four Stokes parameters in the near-ultraviolet spectral region of the Mg II h & k lines over a solar disk area containing part of an active region plage and the edge of a sunspot penumbra. We infer the stratification of the temperature, the electron density, the line of-sight velocity, the micro-turbulent velocity, and the longitudinal component of the magnetic field from the observed intensity and circular polarization profiles. The inferred model atmosphere shows larger temperature and electron density in the plage and the superpenumbra regions than in the quiet regions. The shape of the plage region in terms of its brightness is similar to the pattern of the inferred longitudinal component of the magnetic field in the chromosphere, as well as to that of the overlying moss observed by AIA in the 171 A band, which suggests a similar magnetic origin for the heating in both the plage and the moss region. Moreover, this heating is particularly significant in the regions with larger inferred magnetic flux. In contrast, in the superpenumbra, the regions with larger electron density and temperature are usually found in between these regions with larger magnetic flux, suggesting that the details of the heating mechanism in the chromosphere of the superpenumbra may be different to those in the plage, but with the magnetic field still playing a key role.

astro-ph.SR↗

arXiv2024

Learning Decision Trees and Forests with Algorithmic Recourse

This paper proposes a new algorithm for learning accurate tree-based models while ensuring the existence of recourse actions. Algorithmic Recourse (AR) aims to provide a recourse action for altering the undesired prediction result given by a model. Typical AR methods provide a reasonable action by solving an optimization task of minimizing the required effort among executable actions. In practice, however, such actions do not always exist for models optimized only for predictive performance. To alleviate this issue, we formulate the task of learning an accurate classification tree under the constraint of ensuring the existence of reasonable actions for as many instances as possible. Then, we propose an efficient top-down greedy algorithm by leveraging the adversarial training techniques. We also show that our proposed algorithm can be applied to the random forest, which is known as a popular framework for learning tree ensembles. Experimental results demonstrated that our method successfully provided reasonable actions to more instances than the baselines without significantly degrading accuracy and computational efficiency.

cs.LG↗

arXiv2024

Algorithmic Recourse with Missing Values

This paper proposes a new framework of algorithmic recourse (AR) that works even in the presence of missing values. AR aims to provide a recourse action for altering the undesired prediction result given by a classifier. Existing AR methods assume that we can access complete information on the features of an input instance. However, we often encounter missing values in a given instance (e.g., due to privacy concerns), and previous studies have not discussed such a practical situation. In this paper, we first empirically and theoretically show the risk that a naive approach with a single imputation technique fails to obtain good actions regarding their validity, cost, and features to be changed. To alleviate this risk, we formulate the task of obtaining a valid and low-cost action for a given incomplete instance by incorporating the idea of multiple imputation. Then, we provide some theoretical analyses of our task and propose a practical solution based on mixed-integer linear optimization. Experimental results demonstrated the efficacy of our method in the presence of missing values compared to the baselines.

cs.LG↗

arXiv2024

SCOPE-RL: A Python Library for Offline Reinforcement Learning and Off-Policy Evaluation

This paper introduces SCOPE-RL, a comprehensive open-source Python software designed for offline reinforcement learning (offline RL), off-policy evaluation (OPE), and selection (OPS). Unlike most existing libraries that focus solely on either policy learning or evaluation, SCOPE-RL seamlessly integrates these two key aspects, facilitating flexible and complete implementations of both offline RL and OPE processes. SCOPE-RL put particular emphasis on its OPE modules, offering a range of OPE estimators and robust evaluation-of-OPE protocols. This approach enables more in-depth and reliable OPE compared to other packages. For instance, SCOPE-RL enhances OPE by estimating the entire reward distribution under a policy rather than its mere point-wise expected value. Additionally, SCOPE-RL provides a more thorough evaluation-of-OPE by presenting the risk-return tradeoff in OPE results, extending beyond mere accuracy evaluations in existing OPE literature. SCOPE-RL is designed with user accessibility in mind. Its user-friendly APIs, comprehensive documentation, and a variety of easy-to-follow examples assist researchers and practitioners in efficiently implementing and experimenting with various offline RL methods and OPE estimators, tailored to their specific problem contexts. The documentation of SCOPE-RL is available at https://scope-rl.readthedocs.io/en/latest/.

cs.LG↗

arXiv2024

Towards Assessing and Benchmarking Risk-Return Tradeoff of Off-Policy Evaluation

Off-Policy Evaluation (OPE) aims to assess the effectiveness of counterfactual policies using only offline logged data and is often used to identify the top-k promising policies for deployment in online A/B tests. Existing evaluation metrics for OPE estimators primarily focus on the "accuracy" of OPE or that of downstream policy selection, neglecting risk-return tradeoff in the subsequent online policy deployment. To address this issue, we draw inspiration from portfolio evaluation in finance and develop a new metric, called SharpeRatio@k, which measures the risk-return tradeoff of policy portfolios formed by an OPE estimator under varying online evaluation budgets (k). We validate our metric in two example scenarios, demonstrating its ability to effectively distinguish between low-risk and high-risk estimators and to accurately identify the most efficient one. Efficiency of an estimator is characterized by its capability to form the most advantageous policy portfolios, maximizing returns while minimizing risks during online deployment, a nuance that existing metrics typically overlook. To facilitate a quick, accurate, and consistent evaluation of OPE via SharpeRatio@k, we have also integrated this metric into an open-source software, SCOPE-RL (https://github.com/hakuhodo-technologies/scope-rl). Employing SharpeRatio@k and SCOPE-RL, we conduct comprehensive benchmarking experiments on various estimators and RL tasks, focusing on their risk-return tradeoff. These experiments offer several interesting directions and suggestions for future OPE research.

cs.LG↗

arXiv2024

Determining the nanoflare heating frequency of an X-ray Bright Point observed by MaGIXS

Nanoflares are thought to be one of the prime candidates that can heat the solar corona to its multi-million kelvin temperature. Individual nanoflares are difficult to detect with the present generation instruments, however their presence can be inferred by comparing simulated nanoflare-heated plasma emissions with the observed emission. Using HYDRAD coronal loop simulations, we model the emission from an X-ray bright point (XBP) observed by the Marshall Grazing Incidence X-ray Spectrometer (MaGIXS), along with nearest-available observations from the Atmospheric Imaging Assembly (AIA) onboard Solar Dynamics Observatory (SDO) and X-Ray Telescope (XRT) onboard Hinode observatory. The length and magnetic field strength of the coronal loops are derived from the linear-force-free extrapolation of the observed photospheric magnetogram by Helioseismic and Magnetic Imager (HMI) onboard SDO. Each loop is assumed to be heated by random nanoflares, whose magnitude and frequency are determined by the loop length and magnetic field strength. The simulation results are then compared and matched against the measured intensity from AIA, XRT, and MaGIXS. Our model results indicate the observed emissions from the XBP under study could be well matched by a distribution of nanoflares with average delay times 1500 s to 3000 s, which suggest that the heating is dominated by high-frequency events. Further, we demonstrate the high sensitivity of MaGIXS and XRT to diagnose the heating frequency using this method, while AIA passbands are found to be the least sensitive.

astro-ph.SR↗

arXiv2023

Inverse-Optimization-Based Uncertainty Set for Robust Linear Optimization

We consider solving linear optimization (LO) problems with uncertain objective coefficients. For such problems, we often employ robust optimization (RO) approaches by introducing an uncertainty set for the unknown coefficients. Typical RO approaches require observations or prior knowledge of the unknown coefficient to define an appropriate uncertainty set. However, such information may not always be available in practice. In this study, we propose a novel uncertainty set for robust linear optimization (RLO) problems without prior knowledge of the unknown coefficients. Instead, we assume to have data of known constraint parameters and corresponding optimal solutions. Specifically, we derive an explicit form of the uncertainty set as a polytope by applying techniques of inverse optimization (IO). We prove that the RLO problem with the proposed uncertainty set can be equivalently reformulated as an LO problem. Numerical experiments show that the RO approach with the proposed uncertainty set outperforms classical IO in terms of performance stability.

math.OC↗

arXiv2023

An IPW-based Unbiased Ranking Metric in Two-sided Markets

In modern recommendation systems, unbiased learning-to-rank (LTR) is crucial for prioritizing items from biased implicit user feedback, such as click data. Several techniques, such as Inverse Propensity Weighting (IPW), have been proposed for single-sided markets. However, less attention has been paid to two-sided markets, such as job platforms or dating services, where successful conversions require matching preferences from both users. This paper addresses the complex interaction of biases between users in two-sided markets and proposes a tailored LTR approach. We first present a formulation of feedback mechanisms in two-sided matching platforms and point out that their implicit feedback may include position bias from both user groups. On the basis of this observation, we extend the IPW estimator and propose a new estimator, named two-sided IPW, to address the position bases in two-sided markets. We prove that the proposed estimator satisfies the unbiasedness for the ground-truth ranking metric. We conducted numerical experiments on real-world two-sided platforms and demonstrated the effectiveness of our proposed method in terms of both precision and robustness. Our experiments showed that our method outperformed baselines especially when handling rare items, which are less frequently observed in the training data.

cs.IR↗

arXiv2023

Evidence for the Operation of the Hanle and Magneto-Optical Effects in the Scattering Polarization Signals Observed by CLASP2 Across the Mg II h and k Lines

Radiative transfer investigations of the solar Mg II h and k resonance lines around 280~nm showed that, while their circular polarization (Stokes V) signals arise from the Zeeman effect, the linear polarization profiles (Stokes Q and U) are dominated by the scattering of anisotropic radiation and the Hanle and magneto-optical (MO) effects. Using the unprecedented observations of the Mg II and Mn I resonance lines obtained by the Chromospheric LAyer Spectro-Polarimeter (CLASP2), here we investigate how the linear polarization signals at different wavelengths (i.e., at the center, and at the near and far wings of the k line) vary with the longitudinal component of the magnetic field ($B_{L}$) at their approximate height of formation. The $B_{L}$ is estimated from the V signals in the aforementioned spectral lines. Particular attention is given to the following quantities that are expected to be influenced by the presence of magnetic fields through the Hanle and MO effects: the sign of the U signals, the total linear polarization amplitude ($LP$) and its direction ($χ$) with respect to a reference direction. We find that at the center and near wings of the $k$ line, the behavior of these quantities is significantly different in the observed quiet and plage regions, and that both $LP$ and $χ$ seem to depend on $B_{L}$. These observational results are indicative of the operation of the Hanle effect

astro-ph.SR↗