Gunwoong Park
Seoul National University · 情報科学
研究室紹介
Professor Gunwoong Park's research lab specializes in statistical causal inference and graphical models, with a focus on developing identifiable and computationally efficient algorithms for learning high-dimensional directed acyclic and cyclic graphical models. The lab investigates the theoretical foundations of structural learning in probabilistic models, particularly for discrete and overdispersed data, including Poisson, binomial, and generalized hypergeometric distributions. A central theme is the identification of causal structures from observational data using novel assumptions based on overdispersion and variance-mean relationships.
Research Overview
Research Output Trend
Figures are computed from collected data and may differ slightly.
Selected Papers
15The sense of taste is responsible for the choice of profitable food sources and nestmate recognition in honeybees. Taste detection for food sources occurswithin cuticular hairs located on the antennae, on the mouthparts, and on the tarsi of the forelegs. The gustatory sensilla,which are composed of cuticular hair, respond to sugars, salts, and amino acids. In the honey bee, although sugar detection is a crucial factor in determining the acceptability of nectar and pollen for collection, little i
Olfactory cues are important sensory modalities on individual discrimination, perception, and efficient orientation to food sources in most insects. In honey bees, which are well known as eusocial insects, olfactory cues are mainly used to maintain a colony. Although much research has been reported on olfactory systems in honey bee olfaction, little is known about the differences between two major honey bee species, the European honey bee Apis mellifera and the Asian honey bee Apis cerana. In or
In this paper, we address the question of identifiability and learning algorithms for large-scale Poisson Directed Acyclic Graphical (DAG) models. We define general Poisson DAG models as models where each node is a Poisson random variable with rate parameter depending on the values of the parents in the underlying DAG. First, we prove that Poisson DAG models are identifiable from observational data, and present a polynomial-time algorithm that learns the Poisson DAG model under suitable regulari
Learning DAG or Bayesian network models is an important problem in multi-variate causal inference. However, a number of challenges arises in learning large-scale DAG models including model identifiability and computational complexity since the space of directed graphs is huge. In this paper, we address these issues in a number of steps for a broad class of DAG models where the noise or variance is signal-dependent. Firstly we introduce a new class of identifiable DAG models, where each node has
We introduce a new class of identifiable DAG models where the conditional distribution of each node given its parents belongs to a family of generalized hypergeometric distributions (GHD). A family of generalized hypergeometric distributions includes a lot of discrete distributions such as the binomial, Beta-binomial, negative binomial, Poisson, hyper-Poisson, and many more. We prove that if the data drawn from the new class of DAG models, one can fully identify the graph structure. We further p
Directed graphical models provide a useful framework for modeling causal or directional relationships for multivariate data. Prior work has largely focused on identifiability and search algorithms for directed acyclic graphical (DAG) models. In many applications, feedback naturally arises and directed graphical models that permit cycles occur. In this paper we address the issue of identifiability for general directed cyclic graphical (DCG) models satisfying the Markov assumption. In particular,
Network clustering is a fundamental task that discovers innate communities or groups in networks. Hence, network clustering methods such as spectral clustering and regularized spectral clustering have been applied in a wide range of realms. On top of a network structure, it is known in social network analysis that incorporates information from each vertex can be beneficial. This has led to the development of a series of attributed network clustering algorithms that utilize not only network conne
In this work, we consider the identifiability assumption of Gaussian linear structural equation models (SEMs) in which each variable is determined by a linear function of its parents plus normally distributed error. It has been shown that linear Gaussian structural equation models are fully identifiable if all error variances are the same or known. Hence, this work proves the identifiability of Gaussian SEMs with both homogeneous and heterogeneous unknown error variances. Our new identifiability
We introduce a new class of identifiable DAG models where the conditional distribution of each node given its parents belongs to a family of generalized hypergeometric distributions (GHD). A family of generalized hypergeometric distributions includes a lot of discrete distributions such as the binomial, Beta-binomial, negative binomial, Poisson, hyper-Poisson, and many more. We prove that if the data drawn from the new class of DAG models, one can fully identify the graph structure. We further p
본 연구의 목적은 베이지안 네트워크라고 불리는 방향성 비순환 그래피컬 모델(directed acyclic graphical model; DAG model)을 이용하여 서울시 미세먼지 (PM-10)의 이동 경로를 분석하는 것이다. 이를 위하여 연속형 자료들의 관계망을 찾을 수 있는 정규분포 방향성 비순환 그래피컬 모델 (Gaussian DAG model) 을 이용하였다. 모델의 학습방법으로는 제약 기반 접근법과 점수 기반 접근법을 모두 사용하는 혼합형 max-min hill climbing(MMHC) 알고리즘을 사용하였으며, 이를 통해 미세먼지의 이동 경로가 부분적으로 계절성 풍향과 일치함을 확인하였다. 우리 연구의 결과는 최근 서울시에서 고려중인 공기 정화탑 건설 사업에서 적절한 정화탑 위치를 선정하는데 도움이 될 것으로 기대한다.