• Journal of Korean Society on Water Environment
• /
• v.39 no.1
• /
• pp.1-8
• /
• 2023

Monitoring and prediction of water quality are essential for effective river pollution prevention and water quality management. In this study, a multi-classification model was developed to predict chlorophyll-a (Chl-a) level in rivers. A model was developed using CatBoost, a novel ensemble machine learning algorithm. The model was developed using hourly field monitoring data collected from January 1 to December 31, 2015. For model development, chl-a was classified into class 1 (Chl-a≤10 ㎍/L), class 2 (10<Chl-a≤50 ㎍/L), and class 3 (Chl-a>50 ㎍/L), where the number of data used for the model training were 27,192, 11,031, and 511, respectively. The macro averages of precision, recall, and F1-score for the three classes were 0.58, 0.58, and 0.58, respectively, while the weighted averages were 0.89, 0.90, and 0.89, for precision, recall, and F1-score, respectively. The model showed relatively poor performance for class 3 where the number of observations was much smaller compared to the other two classes. The imbalance of data distribution among the three classes was resolved by using the synthetic minority over-sampling technique (SMOTE) algorithm, where the number of data used for model training was evenly distributed as 26,868 for each class. The model performance was improved with the macro averages of precision, rcall, and F1-score of the three classes as 0.58, 0.70, and 0.59, respectively, while the weighted averages were 0.88, 0.84, and 0.86 after SMOTE application.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.6
• /
• pp.251-256
• /
• 2019

A capsule endoscopy is a medical device that can capture an entire digestive organ from the esophagus to the anus at one time. It produces a vast amount of images consisted of about 8~12 hours in length and more than 50,000 frames on a single examination. However, since the analysis of endoscopic images is performed manually by a medical imaging specialist, the automation requirements of the analysis are increasing to assist diagnosis of the disease in the image. Among them, this study focused on automatic detection of polyp images. A polyp is a protruding lesion that can be found in the gastrointestinal tract. In this paper, we propose a weighted-image generation method to enhance the polyp image learning by multi-scale analysis. It is a way to extract the suspicious region of the polyp through the multi-scale analysis and combine it with the original image to generate a weighted image, that can enhance the polyp image learning. We experimented with SVM and RF which is one of the machine learning methods for 452 pieces of collected data. The F1-score of detecting the polyp with only original images was 89.3%, but when combined with the weighted images generated by the proposed method, the F1-score was improved to about 93.1%.

• Journal of the Korean Statistical Society
• /
• v.29 no.3
• /
• pp.261-268
• /
• 2000

Let {Xn,n$\geq$1} be a sequence of pairwise negative quadrant dependent(NQD) random variables and let {an,n$\geq$1} and {bn,n$\geq$1} be sequencesof constants such that an$\neq$0 and 0$\infty$. In this note, for pairwise NQD random varibles, a general weak law of alrge numbers of the form(∑│aj│Xj-$\upsilon$n)/bnlongrightarrow0) is established, where {νn,n$\geq$1} is a suitable sequence. AMS 2000 subject classifications ; 60F05

• The Pure and Applied Mathematics
• /
• v.20 no.1
• /
• pp.37-50
• /
• 2013

In the theory of hypergeometric functions of one or several variables, a remarkable amount of mathematicians's concern has been given to develop their transformation formulas and summation identities. Here we aim at presenting explicit expressions (in a single form) of the following weighted Appell's function $F_1$: $$(1+2x)^{-a}(1+2z)^{-b}F_1\;\(c,\;a,\;b;\;2c+j;\;\frac{4x}{1+2x},\;\frac{4z}{1+2z}\)\;(j=0,\;{\pm}1,\;{\ldots},\;{\pm}5)$$ in terms of Exton's triple hypergeometric $X_9$. The results are derived with the help of generalizations of Kummer's second theorem very recently provided by Kim et al. A large number of very interesting special cases including Exton's result are also given.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.10
• /
• pp.15-22
• /
• 2020

In this study, we propose an improvement method that can create U-Net model which detect fine concrete cracks by applying a weighted loss function. Because cracks in concrete are a factor that threatens safety, it is important to periodically check the condition and take prompt initial measures. However, currently, the visual inspection is mainly used in which the inspector directly inspects and evaluates with naked eyes. This has limitations not only in terms of accuracy, but also in terms of cost, time and safety. Accordingly, technologies using deep learning is being researched so that minute cracks generated in concrete structures can be detected quickly and accurately. As a result of attempting crack detection using U-Net in this study, it was confirmed that it could not detect minute cracks. Accordingly, as a result of verifying the performance of the model trained by applying the suggested weighted loss function, a highly reliable value (Accuracy) of 99% or higher and a harmonic average (F1_Score) of 89% to 92% was derived. The performance of the learning improvement plan was verified through the results of accurately and clearly detecting cracks.

• Bulletin of the Korean Mathematical Society
• /
• v.56 no.3
• /
• pp.745-756
• /
• 2019

We obtain Lipschitz type characterization and double integral characterization for Fock-type spaces with the norm $${\parallel}f{\parallel}^p_{F^p_{m,{\alpha},t}}\;=\;{\displaystyle\smashmargin{2}{\int\nolimits_{{\mathbb{C}}^n}}\;{\left|{f(z){e^{-{\alpha}}{\mid}z{\mid}^m}}\right|^p}\;{\frac{dV(z)}{(1+{\mid}z{\mid})^t}}$$, where ${\alpha}>0$, $t{\in}{\mathbb{R}}$, and $m{\in}\mathbb{N}$. The results of this paper are the extensions of the classical weighted Fock space $F^p_{2,{\alpha},t}$.

• Communications of the Korean Mathematical Society
• /
• v.32 no.3
• /
• pp.779-787
• /
• 2017

Let m be the Lebesgue measure on ${\mathbb{C}}$ normalized to $m(D)=1,{\mu}$ be an invariant measure on D defined by $d_{\mu}(z)=(1-{\mid}z{\mid}^2)^{-2}dm(z)$. For $f{\in}L^1(D^n,m{\times}{\cdots}{\times}m)$, Bf the Berezin transform of f is defined by, $$(Bf)(z_1,{\ldots},z_n)={\displaystyle\smashmargin{2}{\int\nolimits_D}{\cdots}{\int\nolimits_D}}f({\varphi}_{z_1}(x_1),{\ldots},{\varphi}_{z_n}(x_n))dm(x_1){\cdots}dm(x_n)$$. We prove that if $f{\in}L^1(D^2,{\mu}{\times}{\mu})$ is radial and satisfies ${\int}{\int_{D^2}}fd{\mu}{\times}d{\mu}=0$, then for every bounded radial function ${\ell}$ on $D^2$ we have $$\lim_{n{\rightarrow}{\infty}}{\displaystyle\smashmargin{2}{\int\int\nolimits_{D^2}}}(B^nf)(z,w){\ell}(z,w)d{\mu}(z)d{\mu}(w)=0$$. Then, using the above property we prove n-harmonicity of bounded function which is invariant under the Berezin transform. And we show the same results for the weighted the Berezin transform in the polydisc.

• Proceedings of the KSMRM Conference
• /
• 2001.11a
• /
• pp.113-113
• /
• 2001

Purpose： In the gradient echo EPI the conventional T2＊-weighted image is poor in signal as well distorted by the field inhomogeneity. By acquiring a proton density image in addition to th T2＊-weighted image at the same scan, the fMRI processing can be improved. Method： The central region of the k space is acquired twice at different time points after th RF pulse while acquiring the other regions onc as described in Fig. 1. In Fig. 1 the segment numbers are chronological. Then, we can get two images of different contrast by interleaving th central region in the k space as done in the dua contrast FSE.

• Phonetics and Speech Sciences
• /
• v.6 no.3
• /
• pp.109-119
• /
• 2014

This article represents a first step in a large study aimed at elucidating the relationship between production and perception involved in sound change of /o/ in (Seoul) Korean. In this paper we present the results of a production study and a perception experiment. For the production study we examined vowel production data of 20 young adult speakers, measuring the first and second formants, then conducted a discriminant analysis based on those values. In terms of their F1-F2 values, the distribution of /o/ and /u/ were close, and even overlapping in some circumstances, which is consistent with the literature. This tendency was more apparent among the female speakers than the males. Moreover, with the females' distributions, /o/ was frequently categorized as /u/, suggesting that the direction of the sound change is indeed increasing from /o/ to /u/. Next, to investigate the effects of this proximity on perception, we used the production data of five randomly selected speakers from the production study as stimuli for a perception experiment in which 21 young adult native speakers of (Seoul) Korean performed a vowel identification task and provided a Goodness rating on a 5-point scale. We found that while rates of correctness were high, when these correctness scores were weighted by the Goodness rating, these "weighted correctness" scores were lower in some cases, indicating a degree of confusion in distinguishing between the two vowels.

• Structural Engineering and Mechanics
• /
• v.1 no.1
• /
• pp.47-60
• /
• 1993

The eigen-equation of a wave traveling over repetitive structure is derived directly form the stiffness matrix formulation, in a form which can be used for the case of the cross stiffness submatrix $K_{ab}$ being singular. The weighted adjoint symplectic orthonormality relation is proved first. Then the general method of solution is derived, which can be used either to find all the eigensolutions, or to find the main eigensolutions for large scale problems.