A5100406187- Advanced MIMO Systems Optimization
- Educational Systems and Policies
- Advanced Wireless Network Optimization
- Cooperative Communication and Network Coding
- Educational Research and Pedagogy
- Education and Learning Interventions
- Anomaly Detection Techniques and Applications
- Network Security and Intrusion Detection
- Mechanical Circulatory Support Devices
- Machine Learning in Healthcare
- Statistical Methods and Inference
- Telecommunications and Broadcasting Technologies
- Gene expression and cancer classification
- Time Series Analysis and Forecasting
- Spine and Intervertebral Disc Pathology
- Bioinformatics and Genomic Networks
- Domain Adaptation and Few-Shot Learning
- Chinese history and philosophy
- Shoulder Injury and Treatment
- Cardiac Arrest and Resuscitation
- Cardiac Structural Anomalies and Repair
- Cardiac, Anesthesia and Surgical Outcomes
- Antenna Design and Analysis
- Spinal Fractures and Fixation Techniques
- Health Systems, Economic Evaluations, Quality of Life
Chung-Ang University
2021-2025
Presbyterian Hospital
2024-2025
New York Hospital Queens
2024-2025
Columbia University
2024-2025
NewYork–Presbyterian Hospital
2024-2025
Columbia University Irving Medical Center
2025
Korea University
2012-2024
University of California, Los Angeles
2019-2022
Electronics and Telecommunications Research Institute
2014-2015
Ajou University
2002
Currently available risk prediction methods are limited in their ability to deal with complex, heterogeneous, and longitudinal data such as that primary care records, or multiple competing risks. This paper develops a novel deep learning approach is able successfully address current limitations of standard statistical approaches land marking joint modeling. Our approach, which we call Dynamic-DeepHit, flexibly incorporates the comprising various repeated measurements (rather than only last...
ABSTRACT Background GLP‐1 RAs improve cardiometabolic outcomes in obese, diabetic, and heart failure patients. Data on the safety efficacy of RA advanced with durable LVAD is limited. Objectives To assess Methods We conducted a single‐center retrospective analysis patients support treated RA. Outcomes included related adverse events up to 1 year post initiation. Results Forty were therapy between 2018 2023. At follow‐up, patient's weight was significantly reduced (116 (98–134) vs. 110...
Prediction of nonhome discharge after open reduction internal fixation (ORIF) distal femur fractures may facilitate earlier planning, potentially decreasing costs and improving outcomes. We aim to develop algorithms predicting time ORIF identify features important for model performance. This is a retrospective cohort study adults in the American College Surgeons National Surgical Quality Improvement Program database who underwent between 2010 2019. The primary outcome was discharge,...
ABSTRACT Introduction Durable left ventricular assist devices (LVADs) improve outcomes for advanced heart failure (HF) patients. Vericiguat, which enhances HF by affecting systemic and pulmonary vasculature, may benefit LVAD patients as well. Methods The study aimed to investigate the safety efficacy of Vericiguat in patient on support. This retrospective analysis included supported with who were treated Vericiguat. Safety comprised LVAD‐related hemocompatibility‐related adverse events...
Active Surveillance (AS) for prostate cancer is a management option that continually monitors early disease and considers intervention if progression occurs. A robust method to incorporate "live" updates of risk during follow-up has hitherto been lacking. To address this, we developed deep learning-based individualised longitudinal survival model using Dynamic-DeepHit-Lite (DDHL) learns data-driven distribution time-to-event outcomes. Further refining outputs, used reinforcement learning...
Mobility management is one of the most essential functionalities in mobile networks, providing seamless services for users. performance has been main focuses up to 5G. 3GPP introduced conditional handover (CHO) 5G improve (HO) performance. CHO a well-rounded technique that can solve trade-off between HO failure (HOF) and ping-pong. However, it incur waste radio resources due several extra preparations. Additionally, achieving an optimal solution balances ping-pong user perceived throughput...
Integration of data from multiple omics techniques is becoming increasingly important in biomedical research. Due to non-uniformity and technical limitations platforms, such integrative analyses on omics, which we refer as views, involve learning incomplete observations with various view-missing patterns. This challenging because i) complex interactions within across observed views need be properly addressed for optimal predictive power ii) patterns flexibly integrated. To address...
BackgroundThere remains a lack of accurate and validated outcome-prediction models in total knee arthroplasty (TKA). While machine learning (ML) is powerful predictive tool, determining the proper algorithm to apply across diverse data sets challenging. AutoPrognosis (AP) novel method that uses automated ML framework incorporate best performing stages prognostic modeling into single well-calibrated algorithm. We aimed compare various methods AP performance complications after...
Reverse total shoulder arthroplasty (rTSA) offers tremendous promise for the treatment of complex pathologies beyond scope anatomic but is associated with a higher rate major postoperative complications. We aimed to design and validate machine learning (ML) model predict complications or readmission following rTSA.We retrospectively reviewed California's Office Statewide Health Planning Development database patients who underwent rTSA between 2015 2017. implemented logistic regression (LR),...
The demand and incidence of anatomic total shoulder arthroplasty (aTSA) procedures is projected to increase substantially over the next decade. There a paucity accurate risk prediction models which would be great utility in minimizing morbidity costs associated with major post-operative complications. Machine learning powerful predictive modeling tool has become increasingly popular, especially orthopedics. We aimed build ML model for complications readmission following primary aTSA.A large...
In this paper, we propose an interference avoidance resource allocation scheme based on graph-coloring algorithm to introduce performance gain using spatial reuse in D2D (Device-to-Device) system. By assigning multiple pairs a single resource, from neighboring is inevitable, which leads degradation. Therefore, first the feedback method that can be practically provided by pair. Then, theory, applied efficiently avoid between pairs. Simulation Results represent proposed outperforms...
Chronic diseases evolve slowly throughout a patient's lifetime creating heterogeneous progression patterns that make clinical outcomes remarkably varied across individual patients. A tool capable of identifying temporal phenotypes based on the patients different and would allow clinicians to better forecast disease by recognizing group similar past patients, design treatment guidelines are tailored specific phenotypes. To build such tool, we propose deep learning approach, which refer as...
In device-to-device (D2D) systems, neighboring D2D pairs provide unavoidable interference when a same resource is allocated to multiple pairs. And, this acts as major bottleneck which significantly deteriorates the link quality and prevents from achieving spatial reuse gain. Hence, in paper, we introduce graph-coloring theory propose allocation algorithm for systems order properly avoid dominant Throughout simulations results, show that proposed scheme provides fairly good performance...
In this paper, we study multiple-input single-output downlink cellular systems which jointly design adaptive inter-cell interference cancellation and user scheduling assuming that partial channel state information (CSI) is shared among base stations (BSs). Since the optimal solution requires high complexity, propose a new low complexity algorithm selects best users their beamforming (BF) strategies in terms of maximizing weighted sum rate. To end, first develop simple threshold criterion for...
In this paper, we optimize antenna locations for a distributed system (DAS) with (DA) ports equipped multiple antennas under per-DA port power constraint. Maximum ratio transmission and scaled zero-forcing beamforming are employed single-user multi-user DAS, respectively. Instead of maximizing the cell average ergodic sum rate, focus on lower bound expected signal-to-noise (SNR) single-cell scenario signal-to-leakage (SLR) two-cell to determine locations. For case, optimization SNR criterion...
In this paper, we study an inter-cell interference cancellation (ICIC) scheme for multicell multiple-input single-output downlink systems where each base station (BS) simultaneously supports multiple active users and only partial channel state information (CSI) is shared among BSs. Each BS jointly selects one of following beamforming strategies: egoistic zero-forcing (ZFBF) which applies the ZFBF home altruistic strategy performs both neighboring ICIC. We first derive closed-form expressions...
As of now, a model for predicting the survival patients with out-of-hospital cardiac arrest has not been established. This study aimed to develop identifying predictors over time in during their stay emergency department, using ensemble-based machine learning. A total 26 013 from Korean nationwide registry were enrolled between January 1 and December 31, 2019. Our model, comprising 38 variables, was developed Survival Quilts improve predictive performance. We found that changes important...
이 글은 영문 잡지 The Korea Magazine 소재 제임스 게일(James Scarth Gale)의 한문 고전 英譯 작품의 번역 양상을 분석함으로써 그의 장르에 따른 원리의 一端을 고찰한 것이다.BR 게일의 텍스트에 관한 분석은 꾸준히 수행되고 있다. 특히 그가 작성했던 '번역의 원칙'에 드러나는 도착어 중심의 번역관을 기반으로 평가와 비평이 전개된 바 그러나 원칙은 성경 번역과 긴밀한 관계를 맺고 있어 재고의 여지가 이에 게일이 왕성하게 활동을 했던 시기에 출간된 Magazine을 대상 자료로 삼아 원리를 산문과 운문으로 구분하여 재검토하였다. 그 개략은 다음과 같다.BR 게일은 산문 및 운문 텍스트를 번역할 때 1차적으로 직역주의 원칙을 고수하였다. 인명 지명과 같은 고유명사를 음역하는 한편 원천 텍스트의 일부를 종종 의도적으로 생략하였다. 경우, 텍스트와 전혀 다른 방식으로 행을 설정하거나 일부 구절을 생략하여 번역하기도 하였다. 이처럼 견지하는 변형시켰다. 기존...
Identifying features significantly influencing the target outcome is crucial for understanding complex relationships, reducing computational costs, and improving model generalization in high-dimensional data. While powerful discovering intricate deep learning-based feature selection methods often overlook inherent group structures data, such as gene pathways or categorical variables. Consequently, these may fail to select informative within relevant groups, potentially leading of less...
Abstract Background Cardiomyopathies account for more than half of the cardiovascular disease during peripartum period. In extreme, patients may present with cardiogenic shock (CS) requiring mechanical circulatory support (MCS). The aim this study was to report our experience CS MCS in Methods We a single‐center retrospective analysis all cases involving period that occurred between 2012 and 2023. Results Eleven were included. Median age 33, median BMI 30.4, 73% underwent caesarian‐section...