• 한국멀티미디어학회논문지
• /
• 제23권7호
• /
• pp.883-890
• /
• 2020

In this paper, we propose a variational expectation-maximization algorithm that computes posterior probabilities from Latent Dirichlet Allocation (LDA) model. The algorithm approximates the intractable posterior distribution of a document term matrix generated from a corpus made up by 50 papers. It approximates the posterior by searching the local optima using lower bound of the true posterior distribution. Moreover, it maximizes the lower bound of the log-likelihood of the true posterior by minimizing the relative entropy of the prior and the posterior distribution known as KL-Divergence. The experimental results indicate that documents clustered to image classification and segmentation are correlated at 0.79 while those clustered to object detection and image segmentation are highly correlated at 0.96. The proposed variational inference algorithm performs efficiently and faster than Gibbs sampling at a computational time of 0.029s.

• 한국멀티미디어학회논문지
• /
• 제23권4호
• /
• pp.595-602
• /
• 2020

We propose a generative probabilistic model with Dirichlet prior distribution for topic modeling and text similarity analysis. It assigns a topic and calculates text correlation between documents within a corpus. It also provides posterior probabilities that are assigned to each topic of a document based on the prior distribution in the corpus. We then present a Gibbs sampling algorithm for inference about the posterior distribution and compute text correlation among 50 abstracts from the papers published by IEEE. We also conduct a supervised learning to set a benchmark that justifies the performance of the LDA (Latent Dirichlet Allocation). The experiments show that the accuracy for topic assignment to a certain document is 76% for LDA. The results for supervised learning show the accuracy of 61%, the precision of 93% and the f1-score of 96%. A discussion for experimental results indicates a thorough justification based on probabilities, distributions, evaluation metrics and correlation coefficients with respect to topic assignment.

• Communications for Statistical Applications and Methods
• /
• 제24권5호
• /
• pp.507-518
• /
• 2017

Volatility plays a crucial role in theory and applications of asset pricing, optimal portfolio allocation, and risk management. This paper proposes a combined model of autoregressive moving average (ARFIMA), generalized autoregressive conditional heteroscedasticity (GRACH), and skewed-t error distribution to accommodate important features of volatility data; long memory, heteroscedasticity, and asymmetric error distribution. A fully Bayesian approach is proposed to estimate the parameters of the model simultaneously, which yields parameter estimates satisfying necessary constraints in the model. The approach can be easily implemented using a free and user-friendly software JAGS to generate Markov chain Monte Carlo samples from the joint posterior distribution of the parameters. The method is illustrated by using a daily volatility index from Chicago Board Options Exchange (CBOE). JAGS codes for model specification is provided in the Appendix.

• 응용통계연구
• /
• 제15권2호
• /
• pp.381-393
• /
• 2002

K-모드(modes) 알고리즘과 락(ROCK) 알고리즘은 대규모 범주형 자료에 적용 가능한 데이터 군집화 방법이다. 이 논문에서는 두 알고리즘을 고찰하였으며, 두 알고리즘의 단점을 보완한 개선된 데이터 군집화 알고리즘을 제안하였다. 그리고 실제자료에 제안된 방법을 적용한 모의실험을 실시하여 제안된 방법이 데이터 군집화의 성능을 향상시킬 수 있음을 보였다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제15권2호
• /
• pp.663-669
• /
• 2014

Background: With recent progress in health science administration, a huge amount of data has been collected from thousands of subjects. Statistical and computational techniques are very necessary to understand such data and to make valid scientific conclusions. The purpose of this paper was to develop a statistical probability model and to predict future survival times for male breast cancer patients who were diagnosed in the USA during 1973-2009. Materials and Methods: A random sample of 500 male patients was selected from the Surveillance Epidemiology and End Results (SEER) database. The survival times for the male patients were used to derive the statistical probability model. To measure the goodness of fit tests, the model building criterions: Akaike Information Criteria (AIC), Bayesian Information Criteria (BIC), and Deviance Information Criteria (DIC) were employed. A novel Bayesian method was used to derive the posterior density function for the parameters and the predictive inference for future survival times from the exponentiated Weibull model, assuming that the observed breast cancer survival data follow such type of model. The Markov chain Monte Carlo method was used to determine the inference for the parameters. Results: The summary results of certain demographic and socio-economic variables are reported. It was found that the exponentiated Weibull model fits the male survival data. Statistical inferences of the posterior parameters are presented. Mean predictive survival times, 95% predictive intervals, predictive skewness and kurtosis were obtained. Conclusions: The findings will hopefully be useful in treatment planning, healthcare resource allocation, and may motivate future research on breast cancer related survival issues.

• 대한한의학방제학회지
• /
• 제17권1호
• /
• pp.75-83
• /
• 2009

Objective : The purpose of this study is to indicate of problems in statistical analysis method of "The Korean Journal of oriental Medical Prescription" and we will be proposed the useful application of the statistical analysis method. Methods : In this paper, we were analysed statistical analysis methodology from published journal articles "The Korean Journal of Oriental Medical Prescription" December, year 2000 to December, year 2008. We were investigated of problems in application of structured analysis methods those journal articles that including statistical analysis techniques and analysis methods. Results : 1. A random allocation of the experimental group and control groups are important factors in the planning process of statistical analysis. However, there are less explanation those journal articles. 2. There are no consideration in specimen size that there will be considerate by the level of significance and statistical test. 3. Many article authors were confused between parametric methods and non-parametric methods that they were applied parametric statistical analysis methods although inapplicable sample size. 4. There were applied the parametric methods consists of t-test instead non-parametric methods in the comparison of average intergroup relations. 5. There were less understanding posterior analysis and were confused with t-test. Conclusion : Our goal was to outline the key methods with a brief discussion of problems(statistical analysis methods), avenues for solutions. we recommend authors to use an appropriate statistical analysis methods for obtaining a more cautions results.

• 대한치과교정학회지
• /
• 제52권6호
• /
• pp.399-411
• /
• 2022

Objective: The objective of this randomized clinical trial was to study the skeletal and dental effects of low-level laser therapy (LLLT) along with a miniscrew-assisted expander (Hyrax) after six months of retention. Methods: After sequence generation, concealed allocation, and implementation, 24 female patients were randomly divided (1:1) into two-groups: bone-borne rapid palatal expansion (BBE) without LLLT (n = 12) and BBE with LLLT (n = 12). Eligibility criteria included female patients aged 10-13 years old with bilateral posterior crossbites. Intraoral and extraoral photographs, cone-beam computed tomography images, and digital study models were obtained before expansion and six months after retention. The 7 mm Hyrax appliance was anchored to four palatal mini-screws, which were activated twice daily for 15 days, then locked and kept in place as a retainer. LLLT was performed in the laser group during expansion and retention, according to the guidelines provided. Results: The records of 24 patients were analyzed. According to the post-retention measurements, both groups showed a significant increase in nasal and maxillary widths and total facial height. In the laser group, the Sella-Nasion-Point A and Point A-Nasion-Point B angles and the interpremolar apical distance were significantly increased. Conclusions: Within the limitations of this study, the results suggest that the parameters and protocol of LLLT do not clinically affect the efficiency of BBE in prepubertal and pubertal patients.

• Radiation Oncology Journal
• /
• 제23권1호
• /
• pp.22-31
• /
• 2005

목적 : 전국의 각 병원 방사선종양학과에서 1998년과 1999년도의 2년간 직장암 진단 하에 수술 후 방사선치료를 시행한 환자들의 자료를 분석하여 한국인 직장암 환자의 전체적인 구성과 특성을 파악하고 치료 내용에 대한 현황을 조사하여 국가적인 자료로 활용하고자 하였다. 대상 및 방법 : 대상 환자의 기준은 1998년부터 1999년 사이에 직장 선암의 수술 후 방사선치료를 시작한 환자로서 육안적 잔여 병소 없이 근치적으로 수술이 이루어진 환자를 대상으로 했으며 직장암이외의 다른 암의 병력이 있거나 과거에 골반에 방사선치료를 받은 병력이 있는 환자는 제외하였다. 각 병원별 치료환자 수에 비례하여 해당 병원의 입력 환자수를 정한 후 PCS본부의 무작위 추출 과정을 통하여 입력할 환자를 선정하였다. 선정된 환자는 웹 기반 PCS시스템을 이용하여 각 병원에서 직접 자료를 입력하였다. 결과 : 전국의 19개 병원에서 총 309명의 환자 자료가 입력되었다. 남녀 성비는 59 : 41이었으며 하단연 기준 종양의 위치는 항문연 6 cm 이내가 $46\%$로 가장 많았다. 수술 전 CEA검사는 $79\%$에서 시행되었으며 이 중 $43\%$에서 6 ng/ml 이상인 것으로 나타났다 수술 전 직장내초음파검사는 50명($16\%$)에서만 시행되었으며 CT 등을 이용한 임상적 병기판정은 274명에서 가능하였으며 stage II가 $32\%$, III가 $48\%$를 차지하였다. 수술 후 조직소견에 의한 병리학적 병기는 stage II가 $34\%$, III가 $63\%$였다. 수술의 방법은 복회음부절제수술이 $38\%$, 저위전방절제술이 $59\%$였으며, 5명의 환자에서는 원격전이가 있었으나 원발병소와 함께 절제되었다. 절제연에서 종양세포가 발견된 경우가 13예였으며, 수술 후 항암화학치료를 받은 환자는 전체의 $91\%$였고 $80\%$의 환자는 정맥주사, $9\%$의 환자는 경구항암제를 투여한 것으로 나타났다. 항암제는 5FU와 leucovorine의 조합이 212명($69\%$)으로 가장 많았고 시행횟수는 6회가 140예($45\%$)로 가장 많았다. 환자의 치료자세는 복와위자세가 251예($81.2\%$)로 나타났고, 치료 조사야 수는 박스형 4문조사가 75예($24.3\%$)로 가장 많았으며 3문조사(후방-양측방)가 201예($65.0\%$)로 그 뒤를 이었다. 치료 시 소장을 조사야 외부로 이동시키기 위한 장치나 소변을 참는 등의 조치는 $40.1\%$의 환자에서 시행되었다. 선량의 처방점은 회전중심점이 140예($45.3\%$), 등선량곡선이 123예로 비슷하게 나타났다. 실제 치료된 조사선량은 $180\~7,740$cGy의 분포를 보였으며 목표선량의 $90\%$이상이 투여된 경우가 287예($92.9\%$)였다. 결론 : 전국 각 병원들의 환자를 종합하여 관찰된 내용은 문헌상 권장되는 것과 비슷한 결과를 보였으며 수술의 범위와 항암화학치료의 방법은 병원에 따라 비교적 다양한 형태로 시행되고 있는 것으로 나타났다. 각 병원의 방사선치료 내용은 환자의 상태에 따라 결정되는 탓에 방사선량과 조사야의 선택에 있어 차이가 관찰되었으며 처방된 치료에 대한 환자의 순응도는 $90\%$ 이상으로 높게 나타났다.