• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제12권3호
• /
• pp.125-132
• /
• 2023

본 연구에서는 식물의 영양분 흡수에 따른 식물 성장뿐만 아니라 기공 기능 및 광합성에도 영향을 끼치는 온실의 수증기압차(VPD, Vapor Pressure Deficit)예측을 위한 머신러닝 모델들의 성능을 비교해보았다. VPD 예측을 위해 온실 내·외부 환경요소 및 시계열 데이터의 시간적 요소들과의 상관관계를 확인하고 상관관계가 높은 요소들이 VPD에 어떤 영향을 미치는지 확인하였다. 예측 모델의 성능을 분석하기 전 분석 시계열 데이터의 양(1일, 3일, 7일), 간격(20분, 1시간)이 예측 성능에 미치는 영향을 확인하여 데이터의 양과 간격을 조절하였다. 마지막으로 4개의 머신러닝 예측 모델(XGB Regressor, LGBM Regressor, Random Forest Regressor 등)을 적용하여 모델별 예측 성능을 비교했다. 모델의 예측 결과로 20분 간격의 1일의 데이터를 사용했을 때 LGBM에서 MAE는 0.008, RMSE는 0.011의 가장 높은 예측 성능을 보였다. 또한 20분 후 VPD 예측에 가장 큰 영향을 미치는 요소는 환경적 요인보다는 과거 20분 전의 VPD(VPD_y__71)임을 확인하였다. 본 연구의 결과를 활용하여 VPD 예측을 통해 작물의 생산성을 높이고, 온실의 결로, 병 발생 예방 등이 가능하다. 향후 온실의 환경 데이터 예측뿐만 아니라 더 나아가 생산량 예측, 스마트팜 제어 모델 등 다양한 분야에 활용할 수 있을 것이다.

• 한국콘텐츠학회논문지
• /
• 제21권8호
• /
• pp.662-674
• /
• 2021

대사증후군은 심혈관질환과 밀접한 연관성을 가지며, 최근 대사증후군의 예측을 통한 예방에 관심이 증가하고 있다. 본 연구의 목적은 최근 한국인을 대상으로 한 대사증후군의 발병을 예측하는 논문을 수집, 분석, 종합하여 체계적 문헌고찰을 위한 것이다. 체계적 문헌고찰을 위해 자료검색은 Pubmed, WOS의 해외DB와 DBPia, KISS의 국내DB에서 검색하였으며, 'Metabolic Syndrome', 'predict', 'Korea' 세개의 키워드를 AND 조건으로 2011~2020년에 게재된 논문을 대상으로 검색하였다. 총 560편의 논문이 검색되었고 자료선정기준에 따라 최종 22편의 논문이 선별되었다. 대사증후군 예측에 가장 활용도가 높은 변수는 WHtR(AUC=0.897)이고, 가장 많이 사용된 분석방법은 로지스틱 회귀분석(63.6%), 가장 높은 정확도를 보이는 분석방법은 XGBOOST(AUC=0.879)였다. 또한 한의학적 체질 분류를 적용하는 경우 예측 정확도가 약간 향상되었다. 본 연구 결과를 토대로 한국인의 최적의 대사증후군 예측과 관리를 위한 대규모의 지속적 연구가 수반되어야 할 것으로 생각된다.

• 대한두경부종양학회지
• /
• 제38권1호
• /
• pp.17-24
• /
• 2022

Background/Objectives: This study analyzed the prognostic significance of clinico-pathologic factors including comprehensive nodal factors in parotid gland cancers (PGCs) patients and constructed a survival prediction model for PGCs patients using machine learning techniques. Materials & Methods: A total of 131 PGCs patients were enrolled in the study. Results: There were 19 cases (14.5%) of lymph nodes (LNs) at the lower neck level and 43 cases (32.8%) involved multiple level LNs metastases. There were 2 cases (1.5%) of metastases to the contralateral LNs. Intraparotid LNs metastasis was observed in 6 cases (4.6%) and extranodal extension (ENE) findings were observed in 35 cases (26.7%). Lymphovascular invasion (LVI) and perineural invasion findings were observed in 42 cases (32.1%) and 49 cases (37.4%), respectively. Machine learning prediction models were constructed using clinico-pathologic factors including comprehensive nodal factors and Decision Tree and Stacking model showed the highest accuracy at 74% and 70% for predicting patient's survival. Conclusion: Lower level LNs metastasis and LNR have important prognostic significance for predicting disease recurrence and survival in PGCs patients. These two factors were used as important features for constructing machine learning prediction model. Our machine learning model could predict PGCs patient's survival with a considerable level of accuracy.

• 재활복지공학회논문지
• /
• 제12권1호
• /
• pp.46-52
• /
• 2018

본 연구는 65세 이상 도시거주 노인 107명을 대상으로 일반적 특성, 만성질환 상태, 낙상 관련 의학적 변수, 균형 관련 자신감, 신체적 능력, 우울감을 평가하는 도구를 통해 낙상에 영향을 미치는 관련 주 요인을 찾고자 하였다. 또한 유의한 차이가 있는 변수들 간의 상관관계를 파악하며, 이 중 낙상을 유발하는 데 높은 영향력이 있는 변수를 도출하여 예측력을 알아보았다. 연구 결과, 낙상군에서 요실금, 발의 통증, 하지근력약화, 만성 질환수 및 복용 약물수 빈도수가 비낙상군에 비해 통계적으로 유의하게 높았다. 또한 ABC (Activities-specific Balance Confidence) 총점, BBS (Berg Balance Scale) 총점, SGDS (Short Geriatric Depression Scale) 총점, FRT(Functional Reach Test) 값에서 통계학적으로 유의한 차이가 있었다. 낙상에 영향을 주는 주요인은 ABC 총점으로 점수가 낮을수록 낙상 위험이 높아짐으로써 균형능력에 대한 자기 확신감이 낮을수록 낙상의 가능성이 높아지는 것으로 나타났으며, ABC, SDGS, BBS 척도가 결합하여 적용될 경우 낙상군과 비낙상군을 구분하는 예측력은 70.1%로 높게 나타났다.

• Communications for Statistical Applications and Methods
• /
• 제25권3호
• /
• pp.297-306
• /
• 2018

Cardiovascular disease (CVD) is the leading cause of death worldwide and has a high mortality rate after onset; therefore, the CVD management requires the development of treatment plans and the prediction of prevalence rates. In our study, age, income, education level, marriage status, diabetes, and obesity were identified as risk factors for CVD. Using these 6 factors, we proposed a nomogram based on a $na{\ddot{i}}ve$ Bayesian classifier model for CVD. The attributes for each factor were assigned point values between -100 and 100 by Bayes' theorem, and the negative or positive attributes for CVD were represented to the values. Additionally, the prevalence rate can be calculated even in cases with some missing attribute values. A receiver operation characteristic (ROC) curve and calibration plot verified the nomogram. Consequently, when the attribute values for these risk factors are known, the prevalence rate for CVD can be predicted using the proposed nomogram based on a $na{\ddot{i}}ve$ Bayesian classifier model.

• Journal of Nutrition and Health
• /
• 제55권6호
• /
• pp.684-698
• /
• 2022

Purpose: Recent studies have reported a significant association between skeletal muscle, muscle strength and non-alcoholic fatty liver disease (NAFLD). The effect of nutrient intake on the prediction of skeletal muscle mass and strength or its suggested correlation with metabolic diseases has been primarily reported in healthy individuals. The current study explores the association between energy intake and handgrip strength (HGS) in individuals with NAFLD. Methods: Data were obtained from the Korea National Health and Nutrition Examination Surveys 2016-2018. Data from 12,469 participants were extracted and 1,293 men and 1,401 women aged 20 years and older were included in the analyses of patients with NAFLD. The presence of NAFLD was determined using the hepatic steatosis index. To estimate relative skeletal muscle strength, HGS was measured using a digital dynamometer and calculated by adjusting the body mass index of the dominant arm. Study subjects in the NAFLD and non-NAFLD groups were separately categorized according to quartiles of the calculated HGS. Results: We found that individuals with low (EQ1) energy intake had lower odds of HGS compared to subjects with high (EQ4) energy intake, irrespective of their NAFLD status (p < 0.0001). However, the HGS did not differ based on the level of protein or fat intake ratio. Additionally, the effect of energy intake on HGS was more pronounced in men than in women. Conclusion: Energy intake was associated with the risk of weak HGS in men with NAFLD. The results indicate that energy intake may be a key factor in nutrition care for NAFLD patients with low muscle function.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제11권3호
• /
• pp.107-114
• /
• 2022

본 논문에서는 질병관리청 국민건강영양조사(KNHANES: Korea National Health and Nutrition Examination Survey) 데이터베이스에서 특징선택 방법으로 고혈압을 감지 예측하는 방법을 개선했다. 또한 만성 고혈압과 관련된 다양한 위험 요인을 확인하였다. 본 논문은 3가지로 나누어, 첫째 결측값을 제거하고 Z-변환을 하는 데이터 전처리 단계이다. 다음은 데이터 셋에서 특징선택법을 기반으로 하는 요인분석(FA)을 사용하는 특징선택 단계이며, 특징선택을 기반으로 다중공선형 분석(MC)와 특징중요도(FI)을 비교했다. 마지막으로 예측분석단계에서 고혈압 위험을 감지하고 예측하는데 적용했다. 본 연구에서는 각 분류 모델에 대해 ROC 곡선(AUC) 아래의 평균 표준 오차(MSE), F1 점수 및 면적을 비교한다. 테스트 결과 제안한 MC-FA-RF모델은 80.12% 가장 높은 정확도를 보이고, MSE, f-score, AUC 모델의 경우 각각 0.106, 83.49%의, 85.96% 으로 나타났다. 이러한 결과는 고혈압위험 예측에 대한 제안된 MC-FA-RF 방법이 다른 방법에 비해 우수함을 보이고 있다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권7호
• /
• pp.2613-2617
• /
• 2015

Radiation induced lung injury has long been considered a treatment limiting factor for patients requiring thoracic radiation. This radiation induced lung injury happens early as well as late. Radiation induced lung injury can occur in two phases viz. early (< 6 months) when it is called radiation pneumonitis and late (>6 months) when it is called radiation induced lung fibrosis. There are multiple factors that can be patient, disease or treatment related that predict the incidence and severity of radiation pneumonitis. Radiation induced damage to the type I pneumocytes is the triggering factor to initiate such reactions. Over the years, radiation therapy has witnessed a paradigm shift in radiation planning and delivery and successfully reduced the incidence of lung injury. Radiation pneumonitis is usually a diagnosis of exclusion. Steroids, ACE inhibitors and pentoxyphylline constitute the cornerstone of therapy. Radiation induced lung fibrosis is another challenging aspect. The pathophysiology of radiation fibrosis includes continuing inflammation and microvascular changes due to pro-angiogenic and profibrogenic stimuli resembling those in adult bronchiectasis. General supportive management, mobilization of airway secretions, anti-inflammatory therapy and management of acute exacerbations remains the treatment option. Radiation induced lung injury is an inevitable accompaniment of thoracic radiation.

• Asian Pacific Journal of Cancer Prevention
• /
• 제13권7호
• /
• pp.3089-3097
• /
• 2012

Background and Aims: Vascular endothelial growth factor (VEGF) is a potential prognostic biomarker for patients with resected gastric cancer. However, its role remains controversial. The objective of this study was to conduct a systematic review and meta-analysis of published literature. Methods: Relevant literature was identified using Medline and survival data from published studies were collected following a methodological assessment. Quality assessment of eligible studies and meta-analysis of hazard ratio (HR) were performed to review the correlation of VEGF overexpression with survival and recurrence in patients with gastric cancer. Results: Our meta-analysis included 44 published studies with 4,794 resected patients. VEGF subtype for the prediction of overall survival (OS) included tissue VEGF (HR=2.13, 95% CI 1.71-2.65), circulating VEGF (HR=4.22, 95% CI 2.47-7.18), tissue VEGF-C (HR=2.21, 95% CI 1.58-3.09), tissue VEGF-D (HR=1.73, 95% CI 1.25-2.40). Subgroup analysis showed that HRs of tissue VEGF for OS were, 1.78 (95% CI 0.90-3.51) and 2.31 (95% CI 1.82-2.93) in non-Asians and Asians, respectively. The meta-analysis was also conducted for disease free survival (DFS) and disease specific survival (DSS). Conclusion: Positive expression of tissue VEGF, circulating VEGF, VEGF-C and VEGF-D were all associated with poor prognosis in resected gastric cancer. However, VEGF demonstrated no significant prognostic value for non-Asian populations. Circulating VEGF may be better than tissue VEGF in predicting prognosis.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권10호
• /
• pp.4215-4218
• /
• 2015

Gastric cancer (GC) is one of the most common malignancies in the world. It is the first cause of cancer deaths in both sexes In Iranian population. Circulating insulin-like growth factor-one (IGF-1) levels have been associated for gastric cancer. IGF-1 protein has central roles involved in the regulation of epithelial cell growth, proliferation, transformation, apoptosis and metastasis. Single nucleotide polymorphism in IGF-1 regulatory elements may lead to alter in IGF-1expression level and GC susceptibility. The aim of this study was to investigate the influence of IGF-1 gene polymorphism (rs5742612) on risk of GC and clinicopathological features for the first time in Iranian population. In total, 241 subjects including 100 patients with GC and 141 healthy controls were recruited in our study. Genotypes were analyzed using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) assay with DNA from peripheral blood. The polymorphism was statistically analyzed to investigate the relationship with the risk of GC and clinicopathological properties. Logistic regression analysis revealed that there was no significant association between rs5742612 and the risk of GC. In addition, no significant association between genotypes and clinicopathological features was observed (p value>0.05). The frequencies of the CC, CT, and TT genotypes were 97%, 3%, and 0%, respectively, among the cases, and 97.9%, 2.1%, and 0%, respectively, among the controls. CC genotype was more frequent in cases and controls. The frequencies of C and T alleles were 98.9% and 1.1% in controls and 98.5% and 1.5% in patient respectively. Our results provide the first evidence that this variant is rare in Iranian population and it may not be a powerful genetic predisposing biomarker for prediction GC clinicopathological features in an Iranian population.