• Korean Journal of Acupuncture
• /
• 제33권3호
• /
• pp.102-113
• /
• 2016

Objectives : This study is aimed at developing and discussing the prediction model of blood stasis pattern of traditional Korean medicine(TKM) using machine learning algorithms: multiple logistic regression and decision tree model. Methods : First, we reviewed the blood stasis(BS) questionnaires of Korean, Chinese, and Japanese version to make a integrated BS questionnaire of patient-reported outcomes. Through a human subject research, patients-reported BS symptoms data were acquired. Next, experts decisions of 5 Korean medicine doctor were also acquired, and supervised learning models were developed using multiple logistic regression and decision tree. Results : Integrated BS questionnaire with 24 items was developed. Multiple logistic regression models with accuracy of 0.92(male) and 0.95(female) validated by 10-folds cross-validation were constructed. By decision tree modeling methods, male model with 8 decision node and female model with 6 decision node were made. In the both models, symptoms of 'recent physical trauma', 'chest pain', 'numbness', and 'menstrual disorder(female only)' were considered as important factors. Conclusions : Because machine learning, especially supervised learning, can reveal and suggest important or essential factors among the very various symptoms making up a pattern identification, it can be a very useful tool in researching diagnostics of TKM. With a proper patient-reported outcomes or well-structured database, it can also be applied to a pre-screening solutions of healthcare system in Mibyoung stage.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2005년도 BIOINFO 2005
• /
• pp.319-324
• /
• 2005

An integrated approach for prostate cancer detection using proteomic data is presented. Due to the high-dimensional feature of proteomic data, the discrete wavelet transform (DWT) is used in the first-stage for data reduction as well as noise removal. After the process of DWT, the dimensionality is reduced from 43,556 to 1,599. Thus, each sample of proteomic data can be represented by 1599 wavelet coefficients. In the second stage, a voting method is used to select a common set of wavelet coefficients for all samples together. This produces a 987-dimension subspace of wavelet coefficients. In the third stage, the Autoassociator algorithm reduces the dimensionality from 987 to 400. Finally, the artificial neural network (ANN) is applied on the 400-dimension space for prostate cancer detection. The integrated approach is examined on 9 categories of 2-class experiments, and also 3- and 4-class experiments. All of the experiments were run 10 times of ten-fold cross-validation (i. e. 10 partitions with 100 runs). For 9 categories of 2-class experiments, the average testing accuracies are between 81% and 96%, and the average testing accuracies of 3- and 4-way classifications are 85% and 84%, respectively. The integrated approach achieves exciting results for the early detection and diagnosis of prostate cancer.

• 대한한방내과학회지
• /
• 제39권4호
• /
• pp.699-707
• /
• 2018

Objectives: This study was designed to investigate characteristics of digital infrared thermal imaging (DITI) and quantitative evaluations in patients with burning mouth syndrome (BMS). Methods: We reviewed the clinical records of 38 patients with BMS who visited the Oral Diseases Clinic of Kyung Hee University Korean Medicine Hospital from March 1st, 2018 to June 30th, 2018. The subjects were evaluated with digital infrared thermal imaging (DITI) and for heart rate variability (HRV), unstimulated salivary flow rate (USFR), and the proportion of coated tongue. Results: Most patients showed higher temperatures on the central part of the tongue (T2) than on the middle of the forehead (T1). The patients tended to have a high Low frequency/High frequency (LF/HF) ratio. Statistically significant negative correlations were noted between the age of patients and the temperature of T1 and T2. Statistically significant negative correlations were also observed between the LF/HF ratio and 'T1-T2' values. Conclusions: This study suggests that DITI and HRV are useful for the validation of patients with BMS. Correlations between the result values suggest that sympathetic function acceleration is related to temperature distribution and, ultimately, to symptoms.

• 응용통계연구
• /
• 제19권2호
• /
• pp.319-330
• /
• 2006

현실세계에서 관찰되는 시그널(signal)은 다양한 주파수(frequency)들의 시그널로 혼합되어 있는 경우가 많다. 예를 들어 태양 흑점 자료의 경우 약 11년 주기와 85년 주기로 변동한다는 사실은 널리 알려져 있다. 또한 경제 시계열 자료의 경우는 통상적으로 계절요인(seasonal component), 순환요인(cyclic component) 그리고 장기적인 추세요인(long-term trend)으로 분해하여 분석한다. 이러한 시계열 자료를 구성요소별로 분해하는 것은 오래된 주제중 하나이다. 전통적인 시계열자료 분석기법으로 스펙트럴 분석기법 등이 널리 사용되고 있으나 시계열 자료들이 비정상(nonstationary)일 경우에는 적용하기 어렵다. Huang et. al(1998)은 경험적 모드분해법(empirical mode decomposition)이라고 하는 자료적응적인(data-adaptive) 방법을 제안하였는데, 비정상성(nonstationarity)에 대한 강건성(robustness)으로 여러 분야에 널리 응용되고 있다. 그러나 Huang et. at(1998)은 잡음(error)에 의해 오염된 자료에 대한 구체적인 처리방법은 제시하지 못하고 있다. 본 논문을 통하여 효율적인 잡음제거 방법을 제안하고자 한다.

• International Journal of Vascular Biomedical Engineering
• /
• 제2권2호
• /
• pp.27-32
• /
• 2004

The objective of the present study is to visualize the steady and pulsatile flow fields in a branching model by using a high-resolution PIV system. A bifurcated flow system was built for the experiments in the steady and pulsatile flows. Harvard pulsatile pump was used to generate the pulsatile velocity waveforms. Conifer powder as the tracing particles was added to water to visualize the flow fields. CCD cameras($1K{\times}1K$(high resolution camera) and $640{\times}480$(low resolution camera)) captured two consecutive particle images at once for the image processing of several cross sections on the flow system. The range validation method and the area interpolation method were used to obtain the final velocity vectors with high accuracy. The results of the image processing clearly showed the recirculation zones and the formation of the paired secondary flows from the distal to the apex of the branch flow in the bifurcated model. The results also indicated that the particle velocities at the inner wall moved faster than the velocities at the outer wall due to the inertial force effects and the helical motions generated in the branch flows as the flow proceeded toward the outer wall. Even though the PIV images from the high resolution camera were closer to the simulation results than the images from the low resolution camera at some locations, both results of the PIV experiments from the two cameras generally agreed quite well with the results from the computer simulations. Therefore, instead of using the expensive stereoscopic PIV or 3D PIV system, the three-dimensional flow fields in a bifurcated model could be easily and exactly investigated by this study.

• 한국수자원학회논문집
• /
• 제36권6호
• /
• pp.1025-1033
• /
• 2003

수공구조물의 설계를 비롯하여, 수자원 분야의 기술적 설계의 기초는 수문자료의 처리와 분석에 중심을 두고 있다고 할 수 있다. 수문 자료의 분석방법 중 가장 보편적이면서도 중요한 방법은 자료들의 관계를 도식적으로 규명하는 회귀분석이다. 수위-유량 관계곡선과 같은 수문 자료에 대한 기존의 매개변수적 회귀모형이 갖는 단점은 자료의 특성에 따라, 복수의 회귀식이 산정되거나 동일자료에 대해서도 서로 다른 회귀식이 산정됨으로써 신뢰할 수 있는 회귀곡선을 만들기가 어렵다는 것이다. 이에 비해 주어진 자료에 의해 도출되는 kernel 회귀모형은 자료의 특성과 경향성을 적절히 표현해 줄 수 있는 방법이다. 본 논문에서는 비매개변수적 방법인 kernel 회귀모형을 분석하고, kernel 회귀모형의 중요 인자인 bandwidth의 선택 방법에 따른 kernel 회귀모형의 특성에 대해 비교 분석하였다.

• 한국근적외분광분석학회:학술대회논문집
• /
• 한국근적외분광분석학회 2001년도 NIR-2001
• /
• pp.4112-4112
• /
• 2001

The feasibility of NIR spectroscopy as a quick and nondestructive method for quality control of uniformity of coating thickness of pharmaceutical tablets was investigated. Near infrared spectra of a set of pharmaceutical tablets with varying coating thickness were measured with a diffuse reflectance fiber optic probe connected to a Broker IFS 28/N FT-NIR spectrometer. The challenging issues encountered in this study included: 1. The similarity of the formulation of the core and coating materials, 2. The lack of sufficient calibration samples and 3. The non-linear relationship between the NIR spectral intensity and coating: thickness. A peak at 7184 $cm^{-1}$ was identified that differed for the coating material and the core material when M spectra were collected at 2 $cm^{-1}$ resolution (0.4 nm at 7184 $cm^{-1}$). The study showed that the coating thickness can be analyzed by polynomial fitting of the peak area of the selected peak, while least squares calibration of the same data failed due to the lack of availability of sufficient calibration samples. Samples of coal powder and solid pieces of coal were analyzed by FT-NIR diffuse reflectance spectroscopy with the goal of predicting their ash content, percentage of volatile components, and energy content. The measurements were performed on a Broker Vector 22N spectrometer with a fiber optic probe. A partial least squares model was constructed for each of the parameters of interest for solid and powdered sample forms separately. Calibration models varied in size from 4 to 10 PLS ranks. Correlation coefficients for these models ranged from 86.6 to 95.0%, with root-mean-square errors of cross validation comparable to the corresponding reference measurement methods. The use of FT-NIR diffuse reflectance measurement techniques was found to be a significant improvement over existing measurement methodologies in terms of speed and ease of use, while maintaining the desired accuracy for all parameters and sample forms.(Figure Omitted).

• 한국근적외분광분석학회:학술대회논문집
• /
• 한국근적외분광분석학회 2001년도 NIR-2001
• /
• pp.4102-4102
• /
• 2001

In multivariate analysis, absorbance spectrum is measured over a band of wavelengths. One does not often pay attention to the size of this wavelength band. However, it is desirable that spectrum is measured at only necessary wavelengths as long as the acceptable accuracy of prediction can be met. In this paper, the method of selecting an optimal band of wavelengths based on the loading vector analysis was proposed and applied for determining total protein in human serum using near-infrared transmission spectroscopy and PLSR. Loading vectors in the full spectrum PLSR were used as reference in selecting wavelengths, but only the first loading vector was used since it explains the spectrum best. Absorbance spectra of sera from 97 outpatients were measured at 1530∼1850 nm with an interval of 2 nm. Total protein concentrations of sera were ranged from 5.1 to 7.7 g/㎗. Spectra were measured by Cary 5E spectrophotometer (Varian, Australia). Serum in the 5 mm-pathlength cuvette was put in the sample beam and air in the reference beam. Full spectrum PLSR was applied to determine total protein from sera. Next, the wavelength region of 1672∼1754 nm was selected based on the first loading vector analysis. Standard Error of Cross Validation (SECV) of full spectrum (1530∼l850 nm) PLSR and selected wavelength PLSR (1672∼1754 nm) was respectively 0.28 and 0.27 g/㎗. The prediction accuracy between the two bands was equal. Wavelength selection based on loading vector in PLSR seemed to be simple and robust in comparison to other methods based on correlation plot, regression vector and genetic algorithm. As a reference of wavelength selection for PLSR, the loading vector has the advantage over the correlation plot since the former is based on multivariate model whereas the latter, on univariate model. Wavelength selection by the first loading vector analysis requires shorter computation time than that by genetic algorithm and needs not smoothing.

• 정보과학회 논문지
• /
• 제41권12호
• /
• pp.1050-1057
• /
• 2014

뇌전도 기반의 뇌-컴퓨터 인터페이스는 향후 손 또는 발과 같은 신체를 대체하거나 사용자의 편의성을 제고하는 등의 다양한 목적으로 여러 산업에서 사용이 될 수 있는 기술이다. 본 논문에서는 경험 모드 분해와 고속푸리에 변환을 통해 동작 상상 뇌전도 신호를 분해하고 특징을 추출하는 방법을 제안한다. 뇌전도 신호 분류 과정은 다음과 같이 3단계로 구성된다. 신호 분해에서는 경험모드분해를 이용하여 뇌전도 신호에 대한 내재모드함수를 생성한다. 특징 추출에서는 파워 스펙트럼 밀도를 이용하여 생성된 내재모드함수의 주파수 대역을 확인한 뒤, 뮤파 대역을 포함하고 있는 내재모드함수에 고속푸리에 변환을 적용하여 움직임 상상에 대한 특징을 추출한다. 특징 분류에서는 서포트 벡터 머신을 사용하여 동작 상상 뇌전도 신호에 대한 특징을 분류하고, 10-교차검증을 통해 분류기의 일반화 성능을 추정한다. 제안하는 방법은 다른 방법들과 비교하여 84.50%의 분류 정확도를 보여주었다.

• 인지과학
• /
• 제28권4호
• /
• pp.299-314
• /
• 2017

In-silico 기반의 약물-표적 단백질 연관관계 예측은 신약 탐색 단계에서 매우 중요하다. 그러나 기존의 예측모델은 입력 값이 고정적이며 표적 단백질의 특질 값이 가공된 데이터로 한정됨으로써 예측 모델의 확장성과 유연성이 부족하다. 본 논문에서는 약물-표적 단백질 연관관계를 예측하는 확장 가능한 형태의 머신러닝 모델을 소개한다. 확장 가능한 머신러닝 모델의 핵심 아이디어는 쌍기반의 뉴럴 네트워크로써, 약물과 단백질의 미가공 데이터를 사용하여 특질을 추출하고 특질 값을 각각의 뉴럴 네트워크 레이어에 입력한다. 이 방법은 추가적인 지식없이 자동적으로 약물과 단백질의 특질을 추출한다. 또한 쌍기반 레이어는 특질 값을 풍부한 저차원의 벡터로 향상 시킴으로써 입력 값의 차이로 인한 편향 학습을 방지한다. PubChem BioAssay(PCBA) 데이터 셋에 기반한 5-폴드 교차 검증법을 통하여 제안한 모델의 성능을 평가했으며, 이전의 모델보다 우월한 성능을 보였다.