Earticle

인공지능이 우리 생활의 다양한 곳에 사용되기 시작하였으며, 최근 그 영역 또한 점차 확대되고 있다. 하지만 인 공지능에 대한 교육이 초등학생을 대상으로 이루어지고 있지 않기 때문에 학생들이 인공지능 기술에 대해 어렵게 인식하는 경향이 있다. 이에 본 논문에서는 교육용 프로그래밍 언어와 인공지능 교육 방법을 고찰하고, 인공지능에 대한 교육을 실시함으로써 학생들의 인공지능 기술에 대한 태도의 변화를 살펴보았다. 이를 위해 학생들의 수준에 적절한 블록형 프로그래밍 언어 기반 인공지능 기술에 대한 교육을 실시하였다. 그리고 학생들의 인공지능 기술에 대한 태도를 단일집단 사전사후 검사를 통해 태도의 변화를 살펴보았다. 그 결과 인공지능에 대한 흥미, 인공지능 기술에 대한 접근 가능성, 학교에서 인공지능 기술에 대한 교육의 필요성에 있어 유의미한 향상을 가져왔다.

Artificial intelligence has begun to be used in various parts of our lives, and recently its sphere has been expanding. However, students tend to find it difficult to recognize artificial intelligence technology because education on artificial intelligence is not being conducted on elementary school students. This paper examined the teaching programming language and artificial intelligence teaching methods, and looked at the changes in students' attitudes toward artificial intelligence technology by conducting education on artificial intelligence. To this end, education on block-type programming language-based artificial intelligence technology was provided to students' level. And we looked at students' attitudes toward artificial intelligence technology through a single group pre-postmortem. As a result, it brought about significant improvements in interest in artificial intelligence, possible access to artificial intelligence technology and the need for education on artificial intelligence technology in schools.

Purpose: The purposes of this study were to develop a prediction model for pressure injury using a machine learning algorithm and to integrate it into clinical practice. Methods: This was a retrospective study of tertiary hospitals in Seoul, Korea. It analyzed patients in 12 departments where many pressure injuries occurred, including 8 general wards and 4 intensive care units from January 2018 to May 2022. In total, 182 variables were included in the model development. A pressure injury prediction model was developed using the gradient boosting algorithm, logistic regression, and decision tree methods, and it was compared to the Braden scale. Results: Among the 1,389,660 general ward cases, there were 451 cases of pressure injuries, and among 139,897 intensive care unit cases, there were 297 cases of pressure injuries. Among the tested prediction models, the gradient boosting algorithm showed the highest predictive performance. The area under the receiver operating characteristic curve of the gradient boosting algorithm's pressure injury prediction model in the general ward and intensive care unit was 0.86 (95% confidence interval, 0.83~0.89) and 0.83 (95% confidence interval, 0.79~0.87), respectively. This model was integrated into the electronic health record system to show each patient's probability for pressure injury occurrence, and the risk factors calculated every hour. Conclusion: The prediction model developed using the gradient boosting algorithm exhibited higher performance than the Braden scale. A clinical decision support system that automatically assesses pressure injury risk allows nurses to focus on patients at high risk for pressure injuries without increasing their workload.

Purpose: This study aimed to develop a machine learning (ML)-based predictive model for hospitallength of stay incorporating clinical, nursing, and healthcare system factors to optimize hospitalresource allocation, improve patient-centered care, and enhance nursing workflow efficiency. Methods: This retrospective study analyzed a large dataset of inpatient electronic medical records froma private tertiary hospital. The dataset was used to develop predictive models for long-term versusshort-term hospitalization. The modeling process involved several ML algorithms, and their performancewas evaluated using standard statistical metrics. The most significant predictive variables wereidentified through an analysis of their feature importance. Results: Among the tested models, the Random Forest algorithm exhibited the highest predictive accuracy,demonstrating strong performance in predicting hospital length of stay. Key influencing factorsincluded the number of consultations, postoperative recovery time, duration of stay in the intensive careunit, the use of third-generation antibiotics, and the need for infection isolation. Patients requiringventilator care, intensive care unit admission, and specific powerful antibiotics were more likely toexperience prolonged hospitalization. Additionally, nursing-related factors such as fall risk and pressureulcer risk were significantly correlated with an extended hospital stay. Conclusion: This study demonstrates that ML models can effectively predict hospital length of stay,aiding in hospital resource management, nursing workforce allocation, and patient safety interventions. The integration of predictive analytics into healthcare systems can support early risk assessment,personalized discharge planning, and overall hospital efficiency.

Purpose: This study aimed to develop a predictive model for the early identification of patients at risk of sepsis, using routinely available clinical information and laboratory test results collected during the initial phase of patient care. Methods: This retrospective analysis included electronic medical records of 22,400 adult patients who presented with suspected infection to a tertiary care university hospital in Korea between January 2013 and May 2024. Patients were classified according to Systemic Inflammatory Response Syndrome (score ≥ 2) or Quick Sequential Organ Failure Assessment (score ≥ 2), in combination with sepsis-related International Classification of Diseases, 10th revision codes. Four different machine learning models were trained and validated using five-fold cross-validation. In addition, Shapley additive explanations analysis was performed to interpret the contribution and clinical relevance of key predictive variables. Results: Among the evaluated models, CatBoost demonstrated the strongest predictive performance. Notably, platelet distribution width, alveolar?arterial oxygen difference, procalcitonin, and the arterial/alveolar oxygen ratio consistently emerged as major predictors. Importantly, several variables that did not reach statistical significance in univariate analysis nevertheless contributed substantially to overall model performance, highlighting the importance of complex, multidimensional interactions among clinical factors. Conclusion: These findings indicate that a model based on simple, routinely collected clinical data can achieve high predictive accuracy and strong generalizability. Such a tool may support early clinical decision-making by multidisciplinary teams, including nurses, across diverse real-world care settings. Further prospective studies are warranted to validate its clinical utility and to assess its potential effects on patient outcomes.

Understanding an accurate trip demand by purpose is crucial for short-term regional planning but also long-term regional planning. In traditional approach, information of purpose-oriented trip demand has been derived from public survey in South Korea such as Travel Diary Survey. This type of data acquiring method maybe useful in a sense that it can capture a meaningful sample regarding to entire country, meanwhile it costs a tremendous amount of budget and time. In this research, we want to offer a novel framework for estimating purposeoriented trip demand with dynamic and effective fashions using data fusion in conjunction with Machine learning techniques. With primary results of this concept, this study showed how several state-of-the-art algorithms, including Deep neural network, UMAP, and random forest in conjunction with Genetic algorithm and Tabu-search for optimization, can contribute to this framework. Although tangible results are yet to come, we expect this framework can contribute to resilience planning such as COVID-19.

토양수분은 가뭄 발생과 해소를 매개하는 핵심 인자로서, 대기-지표-지하수로 이어지는 수문순환의 연결 고리 역할을 한다. 이처럼 가뭄 연구 및 분석을 위한 토양수분 자료 관측 센서를 설치하여 측정하고 있으나, 기상 및 통신 장애로 결측이 발생하여 자료 활용에 불편함을 겪고 있다. 본 연구에서는 해남·예산 지역에 설치된 토양수분 모니터링 시스템의 토양수분 결측 자료를 보간하기 위하여 먼저 강수 자료를 보간하고, 강수 특징변수로 머신러닝의 학습자료를 5개로 구축하여 학습 및 평가자료 정확도 결과를 비교·분석하였다. 연구 결과, 지연, 누적, 시계열 특징변수로 구성한 D, E 학습자료 기반 머신러닝 모형이 훈련·검증자료 정확도가 우수하였다. 평가자료 정확도는 D, E 학습자료 기반 XGB 모형이 우수하였으나, E 학습자료 기반 XGB 모형은 다른 조합 대비 더 많은 경우에서 우수한 정확도를 보였다. 따라서 E 학습자료 기반 XGB 모형을 활용하여 해남·예산 지역에 설치된 토양수분 모니터링 시스템의 10·20cm 깊이 토양수분 결측 자료를 보간하는게 적절하다고 판단하였다.

Soil moisture is a key variable governing drought onset and recovery and a critical link in the hydrological cycle connecting the atmosphere, land surface, and groundwater. Missing observations frequently occur in soil moisture monitoring systems due to meteorological and communication failures, limiting data usability. In this study, missing soil moisture data from monitoring systems in Haenam and Yesan were gap-filled by first correcting precipitation data and constructing five machine-learning training datasets using precipitation-based features. Model performance was evaluated using training, validation, and evaluation datasets. Results indicate that Training D and E datasets, incorporating lagged, accumulated, and time-series precipitation features, combined with the XGB algorithm, showed superior performance. The D– and E–XGB combinations also achieved high accuracy in the evaluation dataset, with the E–XGB model outperforming others in more cases. Therefore, the E-dataset XGB model is suitable for gap-filling 10- and 20-cm soil moisture data in the Haenam and Yesan regions.

Predicting traffic accidents is a challenging task because taking into account uncertainty in modeling traffic accidents is not trivial. To address these issues, this article develops a hybrid modeling pipeline combining unsupervised and supervised learning to predict the level of hazardous road sites and explore the causality of accidents by controlling unobserved heterogeneity issues effectively. Traffic accident data for Won-ju province, Korea, from 2020 to 2021, and external factors affecting traffic accidents, such as average travel speed and weather information, are combined based on road links. Through the modeling pipeline, a clustering technique is adopted to capture unobserved heterogeneous information among roads. Since traffic accident data contains a wide variety of categorical and hierarchical features, ensemble methods such as boosting techniques were applied to handle heterogeneity issues among these features. To explore the relationship between the accident and determinant factors, are adopted to interpret the results of machine learning models. Model-agnostic methods, however, generally provide results based on images, this study also added a process that extracts texts from images to overcome compatible issues with existing road safety management systems.

냉동 컨테이너의 고장은 큰 비용의 손실을 야기하지만, 현재 냉동 컨테이너의 알람 체계는 효율성이 떨어진다. 기존에 냉동 시스템의 시뮬레이션 데이터를 활용한 연구는 존재하지만, 냉동 컨테이너의 실제 운영 데이터를 활용한 연구는 부족하다. 이에 본 연구는 실제 냉동 컨테이너 운영 데이터를 활용하여 고장 원인을 분류하였다. 실제 데이터 에서는 데이터 불균형이 발생하였으며 ENN-SMOTE, 클래스 가중치를 둔 Logistic 회귀분석과 본 연구에서 개발한 2-stage 알고리즘을 비교하여 데이터 불균형문제를 해결하였다. 2-stage 알고리즘은 XGboost, LGBoost, DNN을 사용하여 첫 번째 단계에서는 고장 및 정상을 분류하고, 두 번째 단계에서는 고장의 원인을 분류하는 알고리즘이다. 2-stage 알고리즘에서 LGBoost를 사용한 모델이 99.16%의 정확도로 가장 우수하였다. 본 연구는 데이터 불균형을 해결하기 위해 2-stage 알고리즘을 활용한 최종모델을 제안하며 이는 다른 산업에도 활용할 수 있을 것으로 사료된다.

The failure of the reefer container causes a great loss of cost, but the current reefer container alarm system is inefficient. Existing studies using simulation data of refrigeration systems exist, but studies using actual operation data of refrigeration containers are lacking. Therefore, this study classified the causes of failure using actual refrigerated container operation data. Data imbalance occurred in the actual data, and the data imbalance problem was solved by comparing the logistic regression analysis with ENN-SMOTE and class weight with the 2-stage algorithm developed in this study. The 2-stage algorithm uses XGboost, LGBoost, and DNN to classify faults and normalities in the first step, and to classify the causes of faults in the second step. The model using LGBoost in the 2-stage algorithm was the best with 99.16% accuracy. This study proposes a final model using a two-stage algorithm to solve data imbalance, which is thought to be applicable to other industries.

현대사회는 21세기 초반 지식정보사회를 지나 지능정보사회로 바뀌어 가고 있다. 본 연구에서는 지능정보사회 에서 요구되는 학습자의 핵심역량을 신장시키기 위하여 인공지능의 한 분야인 머신러닝을 기반으로 소프트웨어 교육 교수-학습 모형을 개발하였다. 본 모형은 인공지능 자체에 대한 학습의 부담감을 줄이고, 머신러닝을 활용 하여 문제를 해결하는 과정에서 핵심역량을 신장시키는 것에 중점을 두었다. 개발된 모형의 구체적인 단계는 문 제인식 및 분석, 데이터 수집, 데이터 가공 및 선별, ML모델 훈련 및 평가, ML프로그래밍, 적용 및 해결, 공유 및 환류의 7단계로 구성되어 있다. 본 연구에서 개발한 모형을 학생과 학부모를 대상으로 적용한 결과 긍정적인 반응을 얻을 수 있었으며, 이를 통해 머신러닝 기반의 소프트웨어 교육 프로그램의 개발 및 운영에 작은 밑거름 을 제시할 수 있을 것으로 기대한다.

The society we are living in has being changed to the age of the intelligent information society after passing through the knowledge-based information society in the early 21st century. In this study, we have developed the instructional model for software education based on the machine learning which is a field of artificial intelligence( AI) to enhance the core competencies of learners required in the intelligent information society. This model is focusing on enhancing the core competencies through the process of problem-solving as well as reducing the burden of learning about AI itself. The specific stages of the developed model are consisted of seven levels which are ‘Problem Recognition and Analysis’, ‘Data Collection’, ‘Data Processing and Feature Extraction’, ‘ML Model Training and Evaluation’, ‘ML Programming’, ‘Application and Problem Solving’, and ‘Share and Feedback’. As a result of applying the developed model in this study, we were able to observe the positive response about learning from the students and parents. We hope that this research could suggest the future direction of not only the instructional design but also operation of software education program based on machine learning.

본 연구의 목적은 다양한 암 환자에서 PD-1 또는 PD-L1 억제제와 관련된 중증 면역 관련 혈액학적 부작용 (irHAEs)의 위험을 예측할 수 있는 간단한 머신러닝(ML) 모델을 개발하는 것이다. 우리는 대한민국 대학병원의 전자 의무기록 데이터를 기반으로 한 Observational Medical Outcomes Partnership (OMOP) 공통 데이터 모델을 활용하였 다. 중증 irHAEs는 Common Terminology Criteria for Adverse Events (CTCAE) 버전 5.0에 따라 3등급에서 5등급으 로 정의되었다. 예측 모델은 3차 병원(KHMC)의 주요 데이터 세트를 사용하여 개발되었다. 예측 모델은 특징 중요도 값(FIV)을 기반으로 중요한 특성에 집중하여 최소화되었다. 코호트는 397명의 환자가 포함되었다. 테스트된 ML 알 고리즘 중 Random forest가 가장 우수한 예측 성능을 보였으며(AUROC 0.88), FIV의 합이 전체 모델의 50%로 줄어 든 모델까지 전체 모델과 유사한 예측 성능을 나타났다(AUROC 0.83, P>0.05). 최소 예측 모델은 furosemide, oxygen gas, piperacillin/tazobactam, acetylcysteine과 같은 예측 feature 추출되었다. 간결성과 적절한 예측 성능을 고려할 때, 우 리의 ML 최소 예측 모델은 임상 현장에서 쉽게 적용될 수 있으며, 궁극적으로 암 면역 치료를 받는 환자의 약물 안 전성을 향상시킬 수 있을 것이다.

The objective of this study was to develop a simple machine learning (ML) model predictive of the risk for severe immune-related hematological adverse events (irHAEs) associated with PD-1 or PD-L1 inhibitors in patients with various cancers. We utilized the Observational Medical Outcomes Partnership (OMOP) Common Data Model based on electronic medical record data from a university hospital in South Korea. Severe irHAEs were defined as Grades 3 to 5 according to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. The predictive model was developed using the primary dataset from a tertiary hospital (KHMC). The model was simplified by focusing on critical features based on the feature importance values (FIVs). A total of 397 patients were included in the cohort. Among the tested ML algorithms, Random forest showed the best predictive performance (AUROC 0.88), and a parsimonious model, reduced to 50% of the full model's FIVs, demonstrated similar predictive performance to the full model (AUROC 0.83, P > 0.05). The parsimonious model included predictive features such as furosemide, oxygen gas, piperacillin/tazobactam, and acetylcysteine. Considering its simplicity and adequate predictive performance, our parsimonious ML model can be easily implemented in clinical practice and ultimately enhance medication safety in patients treated with cancer immunotherapy.

창정비를 위한 창정비 대상품목, 창정비형태, 창정비원, 요소개발에 필요한 요소개발비 등은 창정비방침에 따라 결정된다. 획득과 운영유지의 효율적인 연계를 위해서는 소요기획 단계부터 창정비방침에 대한 검토가 필요하다. 창정비에 대한 개념을 어떻게 설정하느냐에 따라 설계와 향후 운영유지에 미치는 영향이 크기 때문이다. 본 연구에서는 획득단계 초기 또는 창정비 방침 설정간 활용할 수 있는 머신러닝 예측모델을 제시하였다. 종속변수로서 창정비 수행여부, 창정비 형태(계획정비 또는 고 장정비), 창정비원(군직 또는 외주), 병행정비(군직과 외주), 창정비 요소개발비를 선정하였다. 독립변수로서 장비가격, 전력화 대수, 시스템 단위(전체 또는 구성품), 요소개발 부품수를 선정하였다. 머신러닝에 활용한 통계모형은 로짓(Logit), 프로빗 (Probit), 선형확률모형(LPM)으로 육군에서 분석한 33개의 창정비 분석 자료를 활용하여 학습과 테스트를 수행하였다. 실험 결과 로짓(Logit) 모형의 정확도가 가장 높았으며, 각 종속변수의 정확도는 최대 100%, 평균 65%~93%로 나타났다.

According to the depot maintenance policy, the depot maintenance item, source, type and the development cost are determined. In order to efficiently link acquisition and operation, it is necessary to review the depot maintenance policy from the planning stage. This is because, depending on how the concept of depot maintenance is set, it has a great impact on the design and future operation and maintenance. In this study, a machine learning prediction model that can be used at the beginning of the acquisition stage or during the establishment of the depot maintenance policy was presented. Whether of not depot maintenance is carried out, depot maintenance source, depot maintenance type and development cost for depot maintenance element are dependent variables. The independent variables are the price of equipment, the quantity, the number of components. The statistical model used for machine learning were Logit, Probit, LPM(Linear Probability Model), and 33 depot maintenance analysis data were trained and tested. As a result of the experiment, the accuracy of the Logit model was the highest, and the accuracy of dependent vairables was up to 100%, and the average accuracy was 65%~93%.

본 연구의 목적은 직업상담지원시스템에 탑재하기 위한 직업 추천 모델의 예측 타당도를 분석하고, 예 측 타당성이 높은 항목을 도출하는 데 있다. 이를 위해 기계학습 알고리즘을 활용해 한국직업정보시스템 (KNOW)의 직업특성 설문 문항을 기반으로 직업 추천 항목을 검토하고, 그 성능을 비교 분석하였다. 연 구 목적을 달성하기 위해 1,610명을 대상으로 인적 속성, 흥미, 능력, 일반 업무 관련 변수를 포함한 설문 조사를 실시하였다. 기계학습 알고리즘 분석 결과, 신경망과 그라디언트 부스팅 알고리즘이 가장 높은 성 능을 보였으며, 설문 항목이 직업 추천에 충분한 실용성을 갖추고 있음을 확인하였다. 변수 중요도 분석 을 통해 전공 중분류, 최종 학력, 업무 수행에 필요한 교육 수준, 실외 근무 빈도, 컴퓨터 업무 활동 중요 도 등의 변수가 직업 추천에 중요한 요소로 나타났다. 또한 직업별 학습 모델 성능을 분석하여 분류 정 밀도가 높은 직업과 낮은 직업을 도출하였다. 본 연구는 기계학습 알고리즘을 활용한 직업 추천 시스템 개발에 중요한 기초 자료를 제공하며, 다양한 알고리즘의 비교 분석을 통해 최적의 직업 추천 모델을 도 출하였으며, 이를 통해 구직자의 직업 선택과 경력 개발을 효과적으로 지원할 수 있을 것이다.

The purpose of this study is to analyze the predictive validity of a job recommendation model for use in a career counseling support system and to identify items with high predictive validity. To achieve this, machine learning algorithms were applied to job recommendation items from the Korean Occupational Information System (KNOW), and their performance was compared and analyzed. A survey of 1,610 participants was conducted, including personal attributes, interests, abilities, and general work-related variables. The analysis revealed that the neural network and gradient boosting algorithms performed best, demonstrating that the survey items are practical for job recommendations. Variable importance analysis identified factors such as major classification, education level, job-specific educational requirements, frequency of outdoor work, and the importance of computer-based tasks as key elements in job recommendations. Additionally, model performance for each occupation was analyzed to determine occupations with high and low classification accuracy. This study provides a solid foundation for developing a job recommendation system using machine learning algorithms and identifies the optimal model to support job seekers in their job selection and career development processes.

MOOC의 높은 중도 탈락 문제가 대두됨에 따라서 본 연구는 이를 예측하는 머신러닝 기반의 시간 분할 검증을 적용한 예측모델을 개발하고자 하였다. 이를 위해 edX의 학습 데이터를 사용하여 시간 분할 설계(temporal split design)를 통해 2023년 학습 데이터(N=1,601)로 모델을 훈련하고, 2024년 학습 데이터(N=1,827)로 검증하였다. 모델 평가는 로지스틱 회귀 분석과 랜덤 포레스트 기법을 사용하였다. 탐색적 분석 결과, 학습 중후반기의 퀴즈 수행도 가 가장 중요한 예측 변수로 나타났으며, 데이터 분포가 다른 환경에서도 중도 이탈의 초기 징후를 효과적으로 탐 지하는 것으로 확인되었다. 각 변수의 기여도를 시각화하기 위해 SHAP 분석을 수행한 결과, 퀴즈 응답의 지속성 이 중도 이탈 예측에 가장 중요한 요소로 확인되었다. 이 결과는 검증된 예측 모델을 활용한 조기 경보 시스템이 실제 교육 현장에서 효과적으로 구현될 수 있음을 보여준다.

With the increasing concern over high dropout rates in massive open online courses (MOOCs), the present study endeavored to develop a machine learning–based time-series prediction model to forecast learner attrition. The model was trained on the 2023 dataset (N=1,601) and validated on the 2024 dataset (N=1,827) through a temporal split design, with the utilization of learning log data from edX. The performance of the model was evaluated using logistic regression and random forest techniques. Exploratory analyses revealed that quiz performance during the mid-to-late stages of the course was the most significant predictor of dropout, and the model effectively detected early signs of attrition even in environments with different data distributions. To visualize the contribution of each variable, SHAP analysis was conducted, confirming that the consistency of quiz completion was the most influential factor in predicting dropout. These results demonstrate the efficacy of implementing an early warning system based on a validated prediction model in real educational settings.

본 연구는 고등학교 학생들의 데이터 리터러시를 함양할 수 있는 머신러닝 기반 AI 융합 수업 모형과 수업 설계 원리를 개발하고, 그에 따른 상세 지침을 개발하는 것을 목적으로 하였다. 이를 위해 선행 문헌 연구를 통해 머신러닝을 기반으로 한 수업 모형과 설계 원리 및 상세 지침을 개발하고, 서울 소재 상업계열 특성화고등학교 학생 15명에게 적용하여 실행하였다. 연구 결과 학생들의 데이터 리터러시가 통계적으로 유의미(p< .001)하게 향 상되었으므로 본 연구의 수업 모형이 학습자의 데이터 리터러시 향상에 긍정적인 영향을 주었음을 확인할 수 있었 고, 앞으로 관련 연구로 이어지길 기대한다.

The purpose of this study is to develop a machine learning-based AI convergence class model and class design principles that can foster data literacy in high school students, and to develop detailed guidelines accordingly. We developed a machine learning-based teaching model, design principles, and detailed guidelines through research on prior literature, and applied them to 15 students at a specialized high school in Seoul. As a result of the study, students' data literacy improved statistically significantly (p<.001), so we confirmed that the model of this study has a positive effect on improving learners' data literacy, and it is expected that it will lead to related research in the future.

본 연구는 기계학습법을 활용하여 스포츠센터의 매출액 예측 및 관련 주요 요인을 파악하여 스포츠센터 운영에 필요한 기초자료를 제공하는데 목적을 두고 있다. 연구의 수행을 위해 2019년 1월부터 2022년 12월까지 부산광역시 스포츠센터의 일 매출액 데이터를 수집하였다. 매출이 발생한 날짜의 기상정보를 수집하고 주말여부, 계절, 부산시 일일 코로나 확진자 데이터를 수집하여 설명변수로 투입하였다. 수집된 데이터는 Python ver 3.8로 분석하였으며 선형 회귀 모형, 랜덤포레스트, XGboost 3가지의 기계학습 모형을 사용해 예측력을 교차 검증하였 다. 교차검증 결과 XGboost 모델이 가장 높은 예측력을 보여주었으며 스포츠센터 매출액에 대한 예측을 실시하였 을 때 57.9%의 정확도를 나타냈다. 요인 중요도 분석결과 계절, 날씨 요인이 많이 사용된 것으로 나타나 본 연구 의 결과를 토대로 기상정보와 날짜 등을 고려하여 스포츠센터 경영자들에게 좀 더 정확한 정보를 제공한다면 시 설 운영에 도움을 줄 수 있을 것으로 판단된다.

This study aims to understand the sales forecast of sports centers and the main factors related to the sales forecast by utilizing machine learning method. To conduct the study, collected daily sales data of the Busan Metropolitan City Sports Center from January 2019 to December 2022. Weather information on the date sales occurred was collected, and data on whether it was a weekend, season, and daily COVID-19 confirmed cases in Busan were collected and used as explanatory variables. Cross-validation results showed the XGboost model to have the highest predictive power, with 57.9% accuracy when predicting sports center sales. As a result of the importance analysis of the factors, it was found that the season and weather factors were frequently used. Based on the results of this research, it is judged that the management of the sports center will be useful in providing more accurate information by considering weather information and dates.

This study examines career trajectories among women with career breaks, using data from the 2019 National Survey of Women on Career Breaks (n=1,138). The data underwent preprocessing, including outlier detection, feature scaling, and class imbalance correction with SMOTEENN. Three machine learning models were evaluated, with the Random Forest model achieving the best performance. Key predictors included flexible leave policies, social insurance, remote work options, and job security. The findings highlight the importance of supportive organizational policies in retaining female employees. Future research should explore longitudinal impacts and additional variables like organizational culture.