• Korean Journal of Cognitive Science
• /
• v.20 no.3
• /
• pp.335-355
• /
• 2009

We examined risk judgment and the accuracy of inference based on two kinds of probabilities in a Bayesian inference task: the death probability from a disease (base rates) and the probability of having a disease with positive results in the screening test (posterior probabilities). Risk information were presented in either a probability or a frequency format. In Study 1, we found a numerical format effect for both base rate and posterior probability. Participants rated information as riskier and inferred more accurately in the frequency condition than in the probability condition for both base rate and posterior probability. However, there was no frequency range effect, which suggested that the ranges of frequency format did not influence risk ratings. In order to find out how the analytic thought system influences risk ratings, we compared the ratings of a computation condition and those of a no-computation condition and still found the numerical format effect in computation condition. In Study 2, we examined the numerical format effect and frequency range effect in a high and a low probability condition and found the numerical format effect at each probability level. This result suggests that people feel riskier in the frequency format than in the probability format regardless of the base rates and the posterior probability. We also found a frequency range effect only for the low base rate condition. Our results were discussed in terms of the dual process theories.

• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.77-87
• /
• 2021

With the development of the Internet and personal computers, various and complex attacks begin to emerge. As the attacks become more complex, signature-based detection become difficult. It leads to the research on behavior-based log anomaly detection. Recent work utilizes deep learning to learn the order and it shows good performance. Despite its good performance, it does not provide any explanation for prediction. The lack of explanation can occur difficulty of finding contamination of data or the vulnerability of the model itself. As a result, the users lose their reliability of the model. To address this problem, this work proposes an explainable log anomaly detection system. In this study, log parsing is the first to proceed. Afterward, sequential rules are extracted by Bayesian posterior probability. As a result, the "If condition then results, post-probability" type rule set is extracted. If the sample is matched to the ruleset, it is normal, otherwise, it is an anomaly. We utilize HDFS datasets for the experiment, resulting in F1score 92.7% in test dataset.

• Asian-Australasian Journal of Animal Sciences
• /
• v.33 no.3
• /
• pp.408-415
• /
• 2020

Objective: The aim of the paper was to compare the fit of data derived from daily automatic milking systems (AMS) and monthly test-day records with the use of lactation curves; data was analysed separately for primiparas and multiparas. Methods: The study was carried out on three Polish Holstein-Friesians (PHF) dairy herds. The farms were equipped with an automatic milking system which provided information on milking performance throughout lactation. Once a month cows were also subjected to test-day milkings (method A4). Most studies described in the literature are based on test-day data; therefore, we aimed to compare models based on both test-day and AMS data to determine which mathematical model (Wood or Wilmink) would be the better fit. Results: Results show that lactation curves constructed from data derived from the AMS were better adjusted to the actual milk yield (MY) data regardless of the lactation number and model. Also, we found that the Wilmink model may be a better fit for modelling the lactation curve of PHF cows milked by an AMS as it had the lowest values of Akaike information criterion, Bayesian information criterion, mean square error, the highest coefficient of determination values, and was more accurate in estimating MY than the Wood model. Although both models underestimated peak MY, mean, and total MY, the Wilmink model was closer to the real values. Conclusion: Models of lactation curves may have an economic impact and may be helpful in terms of herd management and decision-making as they assist in forecasting MY at any moment of lactation. Also, data obtained from modelling can help with monitoring milk performance of each cow, diet planning, as well as monitoring the health of the cow.

• Journal of the Korean Society for Library and Information Science
• /
• v.47 no.1
• /
• pp.39-55
• /
• 2013

It is a difficult problem to use keyword retrieval for low-quality Korean document images because these include adjacent characters that are connected. In addition, images that are created from various fonts are likely to be distorted during acquisition. In this paper, we propose and test a keyword retrieval system, using a support vector machine (SVM) for the retrieval of low-quality Korean document images. We propose a keyword retrieval method using an SVM to discriminate the similarity between two word images. We demonstrated that the proposed keyword retrieval method is more effective than the accumulated Optical Character Recognition (OCR)-based searching method. Moreover, using the SVM is better than Bayesian decision or artificial neural network for determining the similarity of two images.

• Journal of the Korean Society for Precision Engineering
• /
• v.19 no.11
• /
• pp.103-111
• /
• 2002

In this study, the natural flaws in welding parts are classified using the signal pattern classification method. The storage digital oscilloscope including FFT function and enveloped waveform generator is used and the signal pattern recognition procedure is made up the digital signal processing, feature extraction, feature selection and classifier design. It is composed with and discussed using the distance classifier that is based on euclidean distance the empirical Bayesian classifier. feature extraction is performed using the class-mean scatter criteria. The signal pattern classification method is applied to the signal pattern recognition of natural flaws.

• Proceedings of the KSME Conference
• /
• 2003.04a
• /
• pp.304-309
• /
• 2003

In this study, natural flaws in welding parts are classified using the signal pattern classification method. The storage digital oscilloscope including FFT function and enveloped waveform generator is used and the signal pattern recognition procedure is made up the digital signal processing, feature extraction, feature selection and classifier design. It is composed with and discussed using the distance classifier that is based on euclidean distance the empirical Bayesian classifier. Feature extraction is performed using the class-mean scatter criteria. The signal pattern classification method is applied to the signal pattern recognition of natural flaws.

• The Bulletin of The Korean Astronomical Society
• /
• v.43 no.1
• /
• pp.62.2-62.2
• /
• 2018

From the gaussian, near scale-invariant density perturbations observed in the CMB to the late time clustering of galaxies, CDM provides a minimal theoretical explanation for a variety of cosmological data. However accepting this explanation, requires that we include within our cosmic ontology a vacuum energy that is ~122 orders of magnitude lower than QM predictions, or alternatively a new scalar field (dark energy) that has negative pressure. Alternatively, modifications to Einstein's General Relativity have been proposed as a model for cosmic acceleration. Recently there have been many works attempting to test for modified gravity using the large scale clustering of galaxies, ISW, cluster abundance, RSD, 21cm observations, and weak lensing. In this work, we compare various modified gravity models using cosmic shear data from the Deep Lens Survey as well as data from CMB, SNe Ia, and BAO. We use the Bayesian Evidence to quantify the comparison robustly, which naturally penalizes complex models with weak data support. In this poster we present our methodology and preliminary constraints on f(R) gravity.

• The Korean Journal of Applied Statistics
• /
• v.16 no.1
• /
• pp.141-150
• /
• 2003

This paper presents the multiple testing method of an autoregressive parameter in stationary AR(1) model using the usual Bayes factor. As prior distributions of parameters in each model, uniform prior and noninformative improper priors are assumed. Posterior probabilities through the usual Bayes factors are used for the model selection. Finally, to check whether these theoretical results are correct, simulated data and real data are analyzed.

• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.365-376
• /
• 2003

In this paper we reviewed a variety of non-Gaussian time series models, and studied the model selection criteria such as AIC and BIC to select proper models. We also considered the likelihood ratio test and applied it to analysis of Polio data set.

• Journal of the Korean Statistical Society
• /
• v.26 no.4
• /
• pp.477-491
• /
• 1997

A new predictive classification rule for assigning future cases into one of two multivariate normal population (with unknown normal mixture model) is considered. The development involves calculation of posterior probability of each possible normal-mixture model via a default Bayesian test criterion, called intrinsic Bayes factor, and suggests predictive distribution for future cases to be classified that accounts for model uncertainty by weighting the effect of each model by its posterior probabiliy. In this paper, our interest is focused on constructing the classification rule that takes care of uncertainty about the types of covariance matrices (homogeneity/heterogeneity) involved in the model. For the constructed rule, a Monte Carlo simulation study demonstrates routine application and notes benefits over traditional predictive calssification rule by Geisser (1982).