arXiv subjects

Kengo Nakamura

Publications and source records attributed to Kengo Nakamura.

10 recordsLinked to original sources

arXiv2026

Linear-Time Exact Computation of Influence Spread on Bounded-Pathwidth Graphs

Given a network and a set of vertices called seeds to initially inject information, influence spread is the expected number of vertices that eventually receive the information under a certain stochastic model of information propagation. Under the commonly used independent cascade model, influence spread is equivalent to the expected number of vertices reachable from the seeds on a directed uncertain graph, and the exact evaluation of influence spread offers many applications, e.g., influence maximization. Although its evaluation is a \#P-hard task, there is an algorithm that can precisely compute the influence spread in $O(mn\omega_p^2\cdot 2^{\omega_p^2})$ time, where $\omega_p$ is the pathwidth of the graph. We improve this by developing an algorithm that computes the influence spread in $O((m+n)\omega_p^2\cdot 2^{\omega_p^2})$ time. This is achieved by identifying the similarities in the repetitive computations in the existing algorithm and sharing them to reduce computation. Although similar refinements have been considered for the probability computation on undirected uncertain graphs, a greater number of similarities must be leveraged for directed graphs to achieve linear time complexity.

cs.DS↗

arXiv2026

Variance Computation for Weighted Model Counting with Knowledge Compilation Approach

One of the most important queries in knowledge compilation is weighted model counting (WMC), which has been applied to probabilistic inference on various models, such as Bayesian networks. In practical situations on inference tasks, the model's parameters have uncertainty because they are often learned from data, and thus we want to compute the degree of uncertainty in the inference outcome. One possible approach is to regard the inference outcome as a random variable by introducing distributions for the parameters and evaluate the variance of the outcome. Unfortunately, the tractability of computing such a variance is hardly known. Motivated by this, we consider the problem of computing the variance of WMC and investigate this problem's tractability. First, we derive a polynomial time algorithm to evaluate the WMC variance when the input is given as a structured d-DNNF. Second, we prove the hardness of this problem for structured DNNFs, d-DNNFs, and FBDDs, which is intriguing because the latter two allow polynomial time WMC algorithms. Finally, we show an application that measures the uncertainty in the inference of Bayesian networks. We empirically show that our algorithm can evaluate the variance of the marginal probability on real-world Bayesian networks and analyze the impact of the variances of parameters on the variance of the marginal.

cs.AI↗

arXiv2025

Tensor Decomposition Meets Knowledge Compilation: A Study Comparing Tensor Trains with OBDDs

A knowledge compilation map analyzes tractable operations in Boolean function representations and compares their succinctness. This enables the selection of appropriate representations for different applications. In the knowledge compilation map, all representation classes are subsets of the negation normal form (NNF). However, Boolean functions may be better expressed by a representation that is different from that of the NNF subsets. In this study, we treat tensor trains as Boolean function representations and analyze their succinctness and tractability. Our study is the first to evaluate the expressiveness of a tensor decomposition method using criteria from knowledge compilation literature. Our main results demonstrate that tensor trains are more succinct than ordered binary decision diagrams (OBDDs) and support the same polytime operations as OBDDs. Our study broadens their application by providing a theoretical link between tensor decomposition and existing NNF subsets.

cs.DS↗

arXiv2024

Single Family Algebra Operation on BDDs and ZDDs Leads To Exponential Blow-Up

Binary decision diagram (BDD) and zero-suppressed binary decision diagram (ZDD) are data structures to represent a family of (sub)sets compactly, and it can be used as succinct indexes for a family of sets. To build BDD/ZDD representing a desired family of sets, there are many transformation operations that take BDDs/ZDDs as inputs and output BDD/ZDD representing the resultant family after performing operations such as set union and intersection. However, except for some basic operations, the worst-time complexity of taking such transformation on BDDs/ZDDs has not been extensively studied, and some contradictory statements about it have arisen in the literature. In this paper, we show that many transformation operations on BDDs/ZDDs, including all operations for families of sets that appear in Knuth's book, cannot be performed in worst-case polynomial time in the size of input BDDs/ZDDs. This refutes some of the folklore circulated in past literature and resolves an open problem raised by Knuth. Our results are stronger in that such blow-up of computational time occurs even when the ordering, which has a significant impact on the efficiency of treating BDDs/ZDDs, is chosen arbitrarily.

cs.DS↗

arXiv2024

Combined Pre-Supernova Alert System with Kamland and Super-Kamiokande

Preceding a core-collapse supernova, various processes produce an increasing amount of neutrinos of all flavors characterized by mounting energies from the interior of massive stars. Among them, the electron antineutrinos are potentially detectable by terrestrial neutrino experiments such as KamLAND and Super-Kamiokande via inverse beta decay interactions. Once these pre-supernova neutrinos are observed, an early warning of the upcoming core-collapse supernova can be provided. In light of this, KamLAND and Super-Kamiokande, both located in the Kamioka mine in Japan, have been monitoring pre-supernova neutrinos since 2015 and 2021, respectively. Recently, we performed a joint study between KamLAND and Super-Kamiokande on pre-supernova neutrino detection. A pre-supernova alert system combining the KamLAND detector and the Super-Kamiokande detector was developed and put into operation, which can provide a supernova alert to the astrophysics community. Fully leveraging the complementary properties of these two detectors, the combined alert is expected to resolve a pre-supernova neutrino signal from a 15 M$_{\odot}$ star within 510 pc of the Earth, at a significance level corresponding to a false alarm rate of no more than 1 per century. For a Betelgeuse-like model with optimistic parameters, it can provide early warnings up to 12 hours in advance.

hep-ex↗

arXiv2023

Tomographic Imaging of the Sagittarius Spiral Arm's Magnetic Field Structure

The Galactic global magnetic field is thought to play a vital role in shaping Galactic structures such as spiral arms and giant molecular clouds. However, our knowledge of magnetic field structures in the Galactic plane at different distances is limited, as measurements used to map the magnetic field are the integrated effect along the line of sight. In this study, we present the first-ever tomographic imaging of magnetic field structures in a Galactic spiral arm. Using optical stellar polarimetry over a $17' \times 10'$ field of view, we probe the Sagittarius spiral arm. Combining these data with stellar distances from the $Gaia$ mission, we can isolate the contributions of five individual clouds along the line of sight by analyzing the polarimetry data as a function of distance. The observed clouds include a foreground cloud ($d < 200$ pc) and four clouds in the Sagittarius arm at 1.23 kpc, 1.47 kpc, 1.63 kpc, and 2.23 kpc. The column densities of these clouds range from 0.5 to $2.8 \times 10^{21}~\mathrm{cm}^{-2}$. The magnetic fields associated with each cloud show smooth spatial distributions within their observed regions on scales smaller than 10 pc and display distinct orientations. The position angles projected on the plane-of-sky, measured from the Galactic north to east, for the clouds in increasing order of distance are $135^\circ$, $46^\circ$, $58^\circ$, $150^\circ$, and $40^\circ$, with uncertainties of a few degrees. Notably, these position angles deviate significantly from the direction parallel to the Galactic plane.

astro-ph.GA↗

arXiv2023

Gamma rays from a reverse shock with turbulent magnetic fields in GRB 180720B

Gamma-ray bursts (GRBs) are the most electromagnetically luminous cosmic explosions. They are powered by collimated streams of plasma (jets) ejected by a newborn stellar-mass black hole or neutron star at relativistic velocities (near the speed of light). Their short-lived (typically tens of seconds) prompt $γ$-ray emission from within the ejecta is followed by long-lived multi-wavelength afterglow emission from the ultra-relativistic forward shock. This shock is driven into the circumburst medium by the GRB ejecta that are in turn decelerated by a mildly-relativistic reverse shock. Forward shock emission was recently detected up to teraelectronvolt-energy $γ$-rays, and such very-high-energy emission was also predicted from the reverse shock. Here we report the detection of optical and gigaelectronvolt-energy $γ$-ray emission from GRB 180720B during the first few hundred seconds, which is explained by synchrotron and inverse-Compton emission from the reverse shock propagating into the ejecta, implying a low-magnetization ejecta. Our optical measurements show a clear transition from the reverse shock to the forward shock driven into the circumburst medium, accompanied by a 90-degree change in the mean polarization angle and fluctuations in the polarization degree and angle. This indicates turbulence with large-scale toroidal and radially-stretched magnetic field structures in the reverse and forward shocks, respectively, which tightly couple to the physics of relativistic shocks and GRB jets -- launching, composition, dissipation and particle acceleration.

astro-ph.HE↗

arXiv2022

Generalization Analysis on Learning with a Concurrent Verifier

Machine learning technologies have been used in a wide range of practical systems. In practical situations, it is natural to expect the input-output pairs of a machine learning model to satisfy some requirements. However, it is difficult to obtain a model that satisfies requirements by just learning from examples. A simple solution is to add a module that checks whether the input-output pairs meet the requirements and then modifies the model's outputs. Such a module, which we call a {\em concurrent verifier} (CV), can give a certification, although how the generalizability of the machine learning model changes using a CV is unclear. This paper gives a generalization analysis of learning with a CV. We analyze how the learnability of a machine learning model changes with a CV and show a condition where we can obtain a guaranteed hypothesis using a verifier only in the inference time. We also show that typical error bounds based on Rademacher complexity will be no larger than that of the original model when using a CV in multi-class classification and structured prediction settings.

cs.LG↗

arXiv2021

Differentiable Equilibrium Computation with Decision Diagrams for Stackelberg Models of Combinatorial Congestion Games

We address Stackelberg models of combinatorial congestion games (CCGs); we aim to optimize the parameters of CCGs so that the selfish behavior of non-atomic players attains desirable equilibria. This model is essential for designing such social infrastructures as traffic and communication networks. Nevertheless, computational approaches to the model have not been thoroughly studied due to two difficulties: (I) bilevel-programming structures and (II) the combinatorial nature of CCGs. We tackle them by carefully combining (I) the idea of \textit{differentiable} optimization and (II) data structures called \textit{zero-suppressed binary decision diagrams} (ZDDs), which can compactly represent sets of combinatorial strategies. Our algorithm numerically approximates the equilibria of CCGs, which we can differentiate with respect to parameters of CCGs by automatic differentiation. With the resulting derivatives, we can apply gradient-based methods to Stackelberg models of CCGs. Our method is tailored to induce Nesterov's acceleration and can fully utilize the empirical compactness of ZDDs. These technical advantages enable us to deal with CCGs with a vast number of combinatorial strategies. Experiments on real-world network design instances demonstrate the practicality of our method.

cs.GT↗

arXiv2020

Variable Shift SDD: A More Succinct Sentential Decision Diagram

The Sentential Decision Diagram (SDD) is a tractable representation of Boolean functions that subsumes the famous Ordered Binary Decision Diagram (OBDD) as a strict subset. SDDs are attracting much attention because they are more succinct than OBDDs, as well as having canonical forms and supporting many useful queries and transformations such as model counting and Apply operation. In this paper, we propose a more succinct variant of SDD named Variable Shift SDD (VS-SDD). The key idea is to create a unique representation for Boolean functions that are equivalent under a specific variable substitution. We show that VS-SDDs are never larger than SDDs and there are cases in which the size of a VS-SDD is exponentially smaller than that of an SDD. Moreover, despite such succinctness, we show that numerous basic operations that are supported in polytime with SDD are also supported in polytime with VS-SDD. Experiments confirm that VS-SDDs are significantly more succinct than SDDs when applied to classical planning instances, where inherent symmetry exists.

cs.DS↗