http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
머신러닝 기반 스마트폰 어플리케이션 멀웨어 감지에 관한 연구
차채연(Chaeyeon Cha),여현영(Hyeonyeong Yeo),심서영(Seoyoung Sim),홍수민(Sumin Hong),박형곤(Hyunggon Park) 한국통신학회 2022 한국통신학회 학술대회논문집 Vol.2022 No.2
본 논문은 스마트폰 어플리케이션이 실행될 때 요구하는 피쳐 데이터를 통해 머신러닝 기반 어플리케이션 멀웨어 여부를 감지하는 모델에 있어 학습에 사용된 데이터의 수와 사용된 머신러닝 알고리즘에 따른 모델의 분류 성능을 분석하였다. 학습에 사용된 데이터의 수와 사용한 알고리즘에 따라 모델의 분류 성능이 달라지는 점을 실제 데이터를 통하여 확인하였다.
노동영,문형곤,한원식,이종원,고은영,김은규,유종한,정소연,여채연,문우경,황기태 한국유방암학회 2010 Journal of breast cancer Vol.13 No.2
Purpose: A better predictive model for occult invasive disease in ductal carcinoma in situ (DCIS) patients is essential to guide the tailored use of sentinel node biopsies. We hypothesized that recent improvement of contrast-enhanced breast magnetic resonance imaging (MRI) could provide more accurate information on the presence of occult invasion in DCIS patients. Methods: From a prospectively maintained database, we identified 143 DCIS patients diagnosed with needle biopsies in whom MRI images were available. Results: Sixty-five patients (45.5%) were upstaged to invasive carcinoma after curative surgery. Ultrasonographic lesion size, mass-appearance on mammography, type of needle used, and the presence of suspicious microinvasive foci were associated with increased likelihood of upstaging. Among the features of MRI, only mass-appearance was significantly associated with the presence of invasive disease (p=0.002). However, up to 50% of masses in MRI cases had massappearance on mammography as well. Other morphologic and pharmacokinetic features of MRI, such as shape, margin, and patterns of enhancement and washout, did not have a significant association. Conclusion: Among various morphologic and pharmacokinetic parameters of contrast-enhanced MRI, only mass-appearance was associated with occult invasive disease. Our results show the limitations of current contrast-enhanced MRI in predicting invasive disease in patients with preoperative diagnoses of DCIS.