• Proceedings of the Safety Management and Science Conference
• /
• 2010.11a
• /
• pp.35-39
• /
• 2010

The research classifies three types of asymptotic tail distributions such as long(heavy, thick) tailed distribution, medium tailed distribution and short(light, thin) tailed distribution. The extreme value distributions(EVD) classified in this paper can be used in SPC(Statistical Process Control) control chart and reliability engineering.

• Journal of the Korea Safety Management & Science
• /
• v.13 no.1
• /
• pp.195-202
• /
• 2011

The study interprets each of three classification models based on Bath-Tub Failure Rate (BTFR), Extreme Value Distribution (EVD) and Conjugate Bayesian Distribution (CBD). The classification model based on BTFR is analyzed by three failure patterns of decreasing, constant, or increasing which utilize systematic management strategies for reliability of time. Distribution model based on BTFR is identified using individual factors for each of three corresponding cases. First, in case of using shape parameter, the distribution based on BTFR is analyzed with a factor of component or part number. In case of using scale parameter, the distribution model based on BTFR is analyzed with a factor of time precision. Meanwhile, in case of using location parameter, the distribution model based on BTFR is analyzed with a factor of guarantee time. The classification model based on EVD is assorted into long-tailed distribution, medium-tailed distribution, and short-tailed distribution by the length of right-tail in distribution, and depended on asymptotic reliability property which signifies skewness and kurtosis of distribution curve. Furthermore, the classification model based on CBD is relied upon conjugate distribution relations between prior function, likelihood function and posterior function for dimension reduction and easy tractability under the occasion of Bayesian posterior updating.

• Annual Conference of KIPS
• /
• 2023.05a
• /
• pp.585-587
• /
• 2023

Long-tailed datasets have an imbalanced distribution because they consist of a different number of data samples for each class. However, there are problems of the performance degradation in tail-classes and class-accuracy imbalance for all classes. To address these problems, this paper suggests a learning method for training of long-tailed dataset. The proposed method uses and combines two methods; one is a resampling method to generate a uniform mini-batch to prevent the performance degradation in tail-classes, and the other is a reweighting method to address the accuracy imbalance problem. The purpose of our proposed method is to train the learning models to have uniform accuracy for each class in a long-tailed dataset.

• Journal of the Korean Mathematical Society
• /
• v.59 no.6
• /
• pp.1185-1201
• /
• 2022

Feller introduced an unfair-fair-game in his famous book [3]. In this game, at each trial, player will win 2^k yuan with probability p_k = 1/2^kk(k + 1), k ∈ ℕ, and zero yuan with probability p₀ = 1 - Σ^∞_k=1 p_k. Because the expected gain is 1, player must pay one yuan as the entrance fee for each trial. Although this game seemed "fair", Feller [2] proved that when the total trial number n is large enough, player will loss n yuan with its probability approximate 1. So it's an "unfair" game. In this paper, we study in depth its convergence in probability, almost sure convergence and convergence in distribution. Furthermore, we try to take 2^k = m to reduce the values of random variables and their corresponding probabilities at the same time, thus a new probability model is introduced, which is called as the related model of Feller's unfair-fair-game. We find out that this new model follows a long-tailed distribution. We obtain its weak law of large numbers, strong law of large numbers and central limit theorem. These results show that their probability limit behaviours of these two models are quite different.

• Journal of Environmental Science International
• /
• v.27 no.12
• /
• pp.1279-1290
• /
• 2018

Sea gulls are high trophic level consumers in the coastal environment, and thus, which have been widely used to monitor contamination biomagnified through a food web. However, such monitoring studies using sea gulls have been rare in the Korean literature. The National Environmental Specimen Bank chose eggs of a black-tailed gulls (Larus crassirostris) to serve as an environmental specimen for the long-term monitoring of the coastal ecosystem affected by terrestrial pollutants. Black-tailed gull eggs were collected from Baengnyeongdo, Hongdo and Uleungdo, and their DDTs and total mercury content were determined. The highest concentration of ${\Sigma}DDTs$ was $231.6{\pm}106.1{\mu}g/kg$ wet in Baengnyeongdo, followed by $230.0{\pm}123.8{\mu}g/kg$ wet in Ulleungdo, and $117.7{\pm}18.3{\mu}g/kg$ wet in Hongdo. In addition, total mercury was detected at $414.5{\pm}97.6{\mu}g/kg$ wet in Ulleungdo, $363.9{\pm}123.6{\mu}g/kg$ wet in Hongdo, and $237.5{\pm}42.3{\mu}g/kg$ wet in Baengnyeongdo. Relatively high concentrations of the target pollutants were recorded in specimens from Ulleungdo. Additional comprehensive and prolonged studies are required to elucidate spatial and temporal patterns of contamination in black-tailed gull eggs with regard to monitoring contaminant trends in eggs and prey.

• Proceedings of the National Institute of Ecology of the Republic of Korea
• /
• v.3 no.4
• /
• pp.191-198
• /
• 2022

Natural habitats of the Korean long-tailed goral (Naemorhedus caudatus) have been fragmented by anthropogenic activities in South Korea in the last decades. Here, the individual identity, genetic variation, and population differentiation of the endangered species were examined via the multiple-tube approach using a non-invasive genotyping method. The average number of alleles was 3.16 alleles/locus for the total population. The Yanggu population (1.66) showed relatively lower average number of alleles than the Inje population (3.67). Of the total 19 alleles, only seven (36.8%) alleles were shared by the two populations. Using five polymorphic out of six loci, four and six different goral individuals from the captive Yanggu (n=24) and the wild Inje (n=28) population were identified, respectively. The allele distribution was not identical between the two populations (Fisher's exact test: P<0.01). A considerably low migration rate was detected between the two populations (no. of migrants after correction for size=0.294). Additionally, the F statistics results indicated significant population differentiation between them, however, quite low (F_ST=0.327, P<0.01). The posterior probabilities indicated that the two populations originated from a single panmictic population (P=0.959) and the assignment test results designated all individuals to both populations with nearly equal likelihood. These could be resulted from moderate population differentiation between the populations. No significant evidence supported recent population bottleneck in the total Korean goral population. This study could provide us with useful population genetic information for conservation and management of the endangered species.

• Communications for Statistical Applications and Methods
• /
• v.12 no.2
• /
• pp.413-423
• /
• 2005

Consider the problem of estimating the underlying regression function from a set of noisy data which is contaminated by a long tailed error distribution. There exist several robust smoothing techniques and these are turned out to be very useful to reduce the influence of outlying observations. However, no matter what kind of robust smoother we use, we should choose the smoothing parameter and relatively less attention has been made for the robust bandwidth selection method. In this paper, we adopt the idea of robust location parameter estimation technique and propose the robust cross validation score functions.

• Journal of the Korean Mathematical Society
• /
• v.54 no.6
• /
• pp.1879-1903
• /
• 2017

In this paper we study the closure property and probability tail asymptotics for randomly weighted sums $S^{\Theta}_n={\Theta}_1X_1+{\cdots}+{\Theta}_nX_n$ for long-tailed random variables $X_1,{\ldots},X_n$ and positive bounded random weights ${\Theta}_1,{\ldots},{\Theta}_n$ under similar dependence structure as in [26]. In particular, we study the case where the distribution of random vector ($X_1,{\ldots},X_n$) is generated by an absolutely continuous copula.

• The Korean Journal of Ecology
• /
• v.22 no.2
• /
• pp.65-68
• /
• 1999

In December-January of 1996-1997 and 1997-1998, information was gathered about vertical distribution of foraging sites of tits in 34 flocks in coniferous and deciduous forests. There was a significant effect of forest type on the distribution of foraging sites of each species. Habitat was classified into 5 height layers vertically: ground, bushes (usually＜1.5 m, up to 3 m), tree layer 1 (up to 1/3 of tree height), tree layer 2 (1/3-2/3 tree height). and tree layer 3 (＞2/3 tree height). There were differences among species: great tit (Parus major) foraged mostly on the ground, coal tit (P. ater) and long-tailed tit (Acrocephalus caudatus) - on the highest tree layer, marsh tit (P. palustris) was often seen on bushes, and varied tit (P. varius) - in tree layer 2. Smaller species used upper and outer parts of trees. suggesting that, like in most other similar studies. larger dominant species prevented smaller species from using inner parts of trees.

• The Korean Journal of Applied Statistics
• /
• v.26 no.6
• /
• pp.987-998
• /
• 2013

This paper considers the statistical characteristics on the air quality (PM10) of Korea collected hourly in 2011. PM10 in Korea exhibits very strong correlations even for higher lags, namely, long range dependence. It is power-law tailed in marginal distribution, and generalized Pareto distribution successfully captures the thicker tail than log-normal distribution. However, slowly decaying autocorrelations may confuse practitioners since a non-stationary model (such as changes in mean) can produce spurious long term correlations for finite samples. We conduct a statistical testing procedure to distinguish two models and argue that the high persistency can be explained by non-stationary changes in mean model rather than long range dependent time series models.