• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제14권3호
• /
• pp.209-215
• /
• 2014

Sequence tagging is the task of predicting frame-wise labels for a given input sequence and has important applications to diverse domains. Conventional methods such as maximum likelihood (ML) learning matches global features in empirical and model distributions, rather than local features, which directly translates into frame-wise prediction errors. Recent probabilistic sequence models such as conditional random fields (CRFs) have achieved great success in a variety of situations. In this paper, we introduce a novel discriminative CRF learning algorithm to minimize local feature mismatches. Unlike overall data fitting originating from global feature matching in ML learning, our approach reduces the total error over all frames in a sequence. We also provide an efficient gradient-based learning method via gradient forward-backward recursion, which requires the same computational complexity as ML learning. For several real-world sequence tagging problems, we empirically demonstrate that the proposed learning algorithm achieves significantly more accurate prediction performance than standard estimators.

• Communications for Statistical Applications and Methods
• /
• 제30권6호
• /
• pp.605-629
• /
• 2023

This paper proposes some diagnostics procedures for the skew-t linear regression model with censored response. The skew-t distribution is an attractive family of asymmetrical heavy-tailed densities that includes the normal, skew-normal and student's-t distributions as special cases. Inspired by the power and wide applicability of the EM-type algorithm, local and global influence analysis, based on the conditional expectation of the complete-data log-likelihood function are developed, following Zhu and Lee's approach. For the local influence analysis, four specific perturbation schemes are discussed. Two real data sets, from education and economics, which are right and left censoring, respectively, are analyzed in order to illustrate the usefulness of the proposed methodology.

• 유통과학연구
• /
• 제16권9호
• /
• pp.43-52
• /
• 2018

Purpose - This article accentuates that there is pervasive corruption in the supply chain and procurement of goods and services in the South African public sector, particularly in the local sphere of government. Consequently, the poor and the indigents who are supposed to receive quality services are affected. Therefore, the article emphasizes the need to strengthen compliance in the supply chain and procurement processes in order to combat corruption. Research design, data, and methodology - This article extensively utilized qualitative literature review approach to address and solve the problem of endemic corruption in the supply chain and procurement being perpetrated by local government officials. Results - The synthesis shows that by strengthening compliance of supply chain and procurement strategies, measures and frameworks, the likelihood of corruption will be drastically reduced if not totally eradicated. It demonstrates that the poor and the indigents who are supposed to receive quality services are being denied due to corruption. It also shows that failure to comply should result to enforcement of punitive sanctions against perpetrators. Conclusions - This study indicates that effective and efficient implementation of supply chain and procurement strategies, measures and frameworks are panacea for curbing corruption in South Africa public sector especially at the local sphere of government.

• 한국인터넷방송통신학회논문지
• /
• 제9권5호
• /
• pp.163-168
• /
• 2009

영상분할의 대부분의 방법들은 각 화소에서 관측되는 특징벡터로 표현하며 이들에 대하여 적절한 확률모델을 가정하게 된다. 이들 확률 모델을 결정하는 파라미터들을 통계적 방법으로 추정하여 이용하거나 각 특징 벡터간의 유사 도를 기반으로 하는 군집 알고리즘을 사용하여 분할을 수행하는 방법들을 이용한다. 이의 대표적인 방법인 EM알고리즘은 불완전한 데이터에서 미지의 파라미터에 대한 최대 우도를 계산하는 경우나 사후 확률 분포의 최대 값을 구하는 문제 등의 응용 분야가 매우 다양하지만 몇 가지의 구조적 문제점을 가지고 있다. 먼저 추정량의 성능이 시작점에 크게 의존한다는 것이며 따라서 우도 함수가 국부적 최대 값에 수렴한다는 것이다. 이러한 문제점을 해결하기 위하여 영상의 모든 레벨 값을 중심으로 형성된 가우시안 함수와 원 영상의 히스토그램을 혼합하여 영상의 새로운 히스토그램을 통해 임계 값을 설정하는 최적화된 영상분할 기법을 제시한다. 제안된 알고리즘은 MFC를 통해 구현하였으며 영상을 임계 값의 개수에 따라 다양하게 나누어 보았을 때 에지부분이 선명하게 나타나며 세밀하고 정확한 영상으로 분할됨을 확인할 수 있다.

• KDI Journal of Economic Policy
• /
• 제42권2호
• /
• pp.31-56
• /
• 2020

This study investigates the impact of the use of the main local languages in Burkina Faso (Moore, Dioula, Fulfulde) on labor force participation. Using Ethnologue language data, I compute the relative language distance reduction index, after which I use a probit/logit model and instrumental variable approach to account for language use policy endogeneity. This study finds that the use of the Moore language increases the likelihood of labor force participation by 36 percent, with a strong impact on women at 59 percent, nine times higher than men, and 38.3 percent for rural individuals, five times higher than individuals living in urban areas. The Dioula language exhibits comparable trends, while Fulfulde has a negative impact on individuals. The study recommends the use of local language(s) as official language(s) to improve labor force participation. However, a bilingual approach combining local and international language(s) will be of use to account for globalization and international competitiveness. The findings here may be of use to researchers and policymakers as part of their effort to increase the labor force participation rates of women and rural individuals. Moreover, this research has significant implications with regard to the implementation of language use policies in a variety of postcolonial language contexts.

• Korean Journal of Otorhinolaryngology-Head and Neck Surgery
• /
• 제61권12호
• /
• pp.702-704
• /
• 2018

The most common cause of treatment failure in oral cavity cancer is when it is found to have local recurrence, usually occurring in the ipsilateral cervical lymph node. On the contrary, it is extremely rare to find local recurrence in soft tissue metastasis (STM) in the contralateral neck. Furthermore, lung cancer and malignant lymphoma are most commonly confined to their primary sites. The poor general condition increases the likelihood of STM, which indicates bad prognosis. A 72-year-old man with a hard and fixed mass on the right submandibular space visited our clinic. He had received a wide excision with local flapreconstruction for squamous cell carcinoma in the left corner of lower lip 18 months ago. We performed the wide excision with bilateral selective neck dissection (I-III), and he was finally diagnosed as STM from contralateral lip cancer. We report this unique and rare disease entity with a literature review.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.162-165
• /
• 2003

Differential learning relies on the differentiated values of nodes, whereas the conventional learning depends on the values themselves of nodes. In this paper, I elucidate the differential learning in the framework maximum likelihood learning of linear generative model with latent variables obeying random walk. I apply the idea of differential learning to the problem independent component analysis(ICA), which leads to differential ICA. Algorithm derivation using the natural gradient and local stability analysis are provided. Usefulness of the algorithm is emphasized in the case of blind separation of temporally correlated sources and is demonstrated through a simple numerical example.

• Journal of the Korean Statistical Society
• /
• 제31권2호
• /
• pp.237-250
• /
• 2002

Motivated by Cook's (1986) assessment of local influence by investigating the curvature of a surface associated with the overall discrepancy measure, this paper extends this idea to the linear regression model with AR(p) disturbances. Diagnostic for the linear regression models with AR(p) disturbances are discussed when simultaneous perturbations of the response vector are allowed. For the derived criterion, numerical studies demonstrate routine application of this work.

• 대한수학회보
• /
• 제54권2호
• /
• pp.573-581
• /
• 2017

Reflected Ornstein-Uhlenbeck process is a process that returns continuously and immediately to the interior of the state space when it attains a certain boundary. In this work, we are concerned with the study of asymptotic behaviours of parametric estimation for ergodic reflected Ornstein-Uhlenbeck processes with two-sided barriers. Moreover, we also focus on the relations between regulators and the local time process.

• Communications for Statistical Applications and Methods
• /
• 제2권2호
• /
• pp.288-295
• /
• 1995

In this paper a hypotheses test procedure based on the least absolute deviation estimators for the unknown parameters in nonlinear regression models is investigated. The asymptotic distribution of the proposed likelihood ratio test statistic are established voth under the null hypotheses and a sequence of local alternative hypotheses. The asymptotic relative efficiency of the proposed test with classical test based on the least squares estimator is also discussed.