http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
[기자회견문] 세계여성폭력추방주간 기념 여성폭력 근절을 위한 여성단체 공동 기자회견_우리는 '생존' 외에 '다른 꿈' 을 꿀 수 있는 세상을 원한다
성매매문제해결을위한전국연대(전국 11개 단체),전국가정폭력상담소협의회(전국 134개소),전국가정폭력피해자보호시설협의회(전국 62개소),전국성폭력상담소협의회(전국 129개소),전국성폭력피해자보호시설협의회(전국20개소),전국이주여성쉼터협의회(전국 19개소),한국성폭력상담소,한국여성단체연합,한국여성의전화(전국 25개 지부),한국이주여성인권센터,UN인권정책센터 한국여신학자협의회 2012 한국여성신학 Vol.- No.76
최적 피처 선택 방법을 적용한 ADHD 분류기의 성능 향상
이덕원,이상협,전국성,김문상 제어로봇시스템학회 2021 제어로봇시스템학회 국내학술대회 논문집 Vol.2021 No.6
In this study, we introduce a method to improve the performance of the ADHD (Attention Deficit Hyperactivity Disorder) classifier by selecting the optimal feature from the data acquired by the Play-type ADHD diagnosis system using robot developed for objective ADHD diagnosis. A total of 84 features are acquired through the ADHD diagnosis system using a robot. However, not all of these features have a good effect on the diagnosis of ADHD in test subjects. Therefore, we selected the optimal features in this study and aimed to improve the performance of the ADHD classifier using only these features. In order to select the optimal feature, we used a Recursive Feature Elimination (RFE) method that can define the priority of features using a simple model such as a linear model. In addition, ADHD was classified by applying the data of the selected feature to the Multi-Layer Perceptron (MLP) and Support Vector Machine (SVM) algorithms. Finally, when 25 features were used, a classification accuracy of 96.32% was achieved compared to that of a specialist.