This study develops a machine learning-based tool life prediction model using spindle power data collected from real manufacturing environments. The primary objective is to monitor tool wear and predict optimal replacement times, thereby enhancing manufacturing efficiency and product quality in smart factory settings. Accurate tool life prediction is critical for reducing downtime, minimizing costs, and maintaining consistent product standards. Six machine learning models, including Random Forest, Decision Tree, Support Vector Regressor, Linear Regression, XGBoost, and LightGBM, were evaluated for their predictive performance. Among these, the Random Forest Regressor demonstrated the highest accuracy with R2 value of 0.92, making it the most suitable for tool wear prediction. Linear Regression also provided detailed insights into the relationship between tool usage and spindle power, offering a practical alternative for precise predictions in scenarios with consistent data patterns. The results highlight the potential for real-time monitoring and predictive maintenance, significantly reducing downtime, optimizing tool usage, and improving operational efficiency. Challenges such as data variability, real-world noise, and model generalizability across diverse processes remain areas for future exploration. This work contributes to advancing smart manufacturing by integrating data-driven approaches into operational workflows and enabling sustainable, cost-effective production environments.

The purpose of this study is to develop a timely fall detection system aimed at improving elderly care, reducing injury risks, and promoting greater independence among older adults. Falls are a leading cause of severe complications, long-term disabilities, and even mortality in the aging population, making their detection and prevention a crucial area of public health focus. This research introduces an innovative fall detection approach by leveraging Mediapipe, a state-of-the-art computer vision tool designed for human posture tracking. By analyzing the velocity of keypoints derived from human movement data, the system is able to detect abrupt changes in motion patterns, which are indicative of potential falls. To enhance the accuracy and robustness of fall detection, this system integrates an LSTM (Long Short-Term Memory) model specifically optimized for time-series data analysis. LSTM's ability to capture critical temporal shifts in movement patterns ensures the system's reliability in distinguishing falls from other types of motion. The combination of Mediapipe and LSTM provides a highly accurate and robust monitoring system with a significantly reduced false-positive rate, making it suitable for real-world elderly care environments. Experimental results demonstrated the efficacy of the proposed system, achieving an F1 score of 0.934, with a precision of 0.935 and a recall of 0.932. These findings highlight the system's capability to handle complex motion data effectively while maintaining high accuracy and reliability. The proposed method represents a technological advancement in fall detection systems, with promising potential for implementation in elderly monitoring systems. By improving safety and quality of life for older adults, this research contributes meaningfully to advancements in elderly care technology.

This study proposes a method to evaluate the publicity of real-time, demand-responsive, autonomous public-transportation systems. By analyzing real-time data collected based on publicity evaluation indicators suggested in previous research studies, this study seeks to establish a system that objectively assesses the publicity of public transportation. Thus, the introduction of autonomous public transportation systems is expected to contribute to solving problems in underserved transportation areas and enable more sophisticated public transportation operations. We reviewed evaluation indicators proposed in previous studies. Based on this review, publicity evaluation indicators were derived and specific criteria were selected to assess systematically the publicity of autonomous public transportation. An AHP analysis was conducted to assess the relative importance of each indicator by analyzing the importance of the selected indicators. Additionally, to score the indicators, minimum and maximum target values were established, and a method for assigning scores to each indicator was examined. The most important factor in the publicity evaluation of autonomous demand-responsive transport (DRT) was the “success rate of allocation to weak public transportation service areas,” with a significance level p of 0.204. This was analyzed as a key evaluation criterion because of the importance of service provision in areas with low-public-transportation accessibility. Subsequently, “Accessing distance to a virtual station” (p = 0.145) was evaluated as an important factor representing the convenience of the service. “Waiting time after allocation” (p = 0.134) also appeared as an important evaluation factor, as reducing waiting time considerably affected service quality. Conversely, “compliance rate of velocity” yielded the lowest significance (p = 0.017), as speed compliance was typically guaranteed owing to autonomous driving technology. This study proposed a specific evaluation method based on publicity indicators to provide a strategic direction for improving services and enhancing the publicity of autonomous DRT systems. These results can serve as a foundational resource for improving transportation services in underserved areas and for enhancing the overall quality of public transportation services. However, the study’s limitation was its inability to use real-time autonomous public transportation data, relying instead on I-MoD data from Incheon. This limitation constrained the ability to establish universal benchmarks because data from various municipalities were not included. Future research should collect and analyze data from diverse regions to establish more reliable evaluation indicators.

This study investigates using Conditional Tabular Generative Adversarial Networks (CT-GAN) to generate synthetic data for turnover prediction in large employment datasets. The effectiveness of CT-GAN is compared with Adaptive Synthetic Sampling (ADASYN), Synthetic Minority Over-sampling Technique (SMOTE), and Random Oversampling (ROS) using Logistic Regression (LR), Linear Discriminant Analysis (LDA), Random Forest (RF), and Extreme Learning Machines (ELM), evaluated with AUC and F1-scores. Results show that GAN-based techniques, especially CT-GAN, outperform traditional methods in addressing data imbalance, highlighting the need for advanced oversampling methods to improve classification accuracy in imbalanced datasets.

This study integrates TabTransformer and CTGAN for predicting job satisfaction among South Korean college graduates. TabTransformer handles complex tabular data relationships with self-attention, while CTGAN generates high-quality synthetic samples. The combined approach achieves an accuracy of 0.85, precision of 0.83, recall of 0.82, F1-score of 0.82, and an AUC of 0.88. Cross-validation confirms the model's robustness and generalizability with a mean accuracy of 0.85 and a standard deviation of 0.008. The integration of TabTransformer and CTGAN enhances predictive accuracy and model generalizability, providing valuable insights for employment policy and research.

해양 환경에서 발생하는 화재는 일반적인 화재 상황에 비해 빠르게 화염이 전파되기 때문에 초기 발견과 대응이 매우 중 요하다. 최근의 화재 감지 시스템은 카메라 센서와 딥러닝 검출 모델을 활용하여 개발되고 있지만, 해양 환경에 특화된 딥러닝 모델 을 학습하기 위해 해양 환경에서 화재 데이터를 실제로 수집하는 것은 기술적, 경제적 측면에서 어려움이 존재한다. 본 논문에서는 이러한 문제를 해결하기 위해 언리얼 엔진 기반 가상 데이터 생성 도구를 활용하여 가상 환경에서 해양 환경을 구축하고 여러 상황 의 시나리오에서 데이터를 수집하여 해양 환경 화재 가상 데이터셋을 구축하였다. 가상 데이터셋으로 학습한 RT-DETR-L 모델은 실 제 해양 환경에서 발생한 화재 상황을 수집하여 제작한 테스트 데이터셋에서 mAP50:95 0.529를 달성하였다. 또한 가상 데이터로 학습 한 검출 모델은 일반적인 화재 상황이나 항만시설에서 연기만 발생하는 상황에서도 화재를 검출하는 것을 볼 수 있었다. 이를 통해 실제 데이터가 아닌 가상 데이터셋을 사용하여 데이터셋을 구축하여도 해양 환경 화재와 같은 특수한 상황에서의 검출 모델 성능 향 상에 도움을 줄 수 있다는 것을 확인하였다.

본 연구는 북한이 2024년을 전쟁 준비 완성의 해로 선언하고 연이어 미사일을 발사하여 안보를 위협하는 상황에서, 빅데이터 분석을 활용하 여 한국 언론보도와 포털 사이트에 나타난 북핵 및 미사일 위협에 대한 담론과 인식의 특성을 실증적으로 분석하고, 그에 따른 시사점을 도출하 는 것을 목적으로 한다. 이를 위해 국내 주요 언론보도와 포털 사이트에 서 총 33,318건의 데이터를 수집하여, TF-IDF 분석을 통해 상위 50개 의 주요 키워드를 도출하고, 사회연결망 분석을 통해 각 키워드 간의 연 결 정도와 구조를 파악하였다. 분석 결과, 러시아-우크라이나 전쟁, 이스 라엘-하마스 전쟁 등 국제적 안보 불안과 동북아에서의 북-러 군사협력 및 한-미-일 군사협력의 대립 구도 등이 사회적 담론 형성에 큰 영향을 미친 것으로 나타났다. 이에 따라 한-미-일 군사협력 강화와 확장 억제 전략의 신뢰성을 높이고, 사회적 차원에서 위기의식과 안보의식의 제고 가 필요하다는 시사점이 도출되었다.