일 | 월 | 화 | 수 | 목 | 금 | 토 |
---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | |||
5 | 6 | 7 | 8 | 9 | 10 | 11 |
12 | 13 | 14 | 15 | 16 | 17 | 18 |
19 | 20 | 21 | 22 | 23 | 24 | 25 |
26 | 27 | 28 | 29 | 30 | 31 |
Tags
- svdd
- SVM 이란
- cs231n lecture5
- Deep Learning
- pytorch project
- computervision
- yolov3
- darknet
- RCNN
- Object Detection
- CNN
- 논문분석
- EfficientNet
- SVM hard margin
- DeepLearning
- TCP
- fast r-cnn
- pytorch
- libtorch
- Faster R-CNN
- CS231n
- support vector machine 리뷰
- self-supervision
- yolo
- Computer Vision
- 데이터 전처리
- cnn 역사
- pytorch c++
- 서포트벡터머신이란
- SVM margin
Archives
- Today
- Total
목록distance based anomaly detection (1)
아롱이 탐험대
[Anomaly Detection] Distance, Clustering, PCA-based Methods
본 포스트는 고려대학교 산업경영공학부 강필성 교수님의 Business Analytics 강의를 정리한 내용입니다. 1. Distance-based anomaly detection Distance 기반 anomaly detection은 outlier에 해당하는 객체들은 이웃들과의 거리가 멀 것이라고 가정한다. 이 말은 즉슨 정상 데이터는 근처에 다른 데이터들이 많을 것이라는 것이다. 여기서 핵심은 Normal class에 대해 확률 분포를 가정하지 않는다. 오직 객체 간의 거리만 계산해서 anomaly socre를 측정한다. KNN based Anomaly Detection을 알아보기 전 KNN에 대해 알아보자. KNN traing set에서 주어진 test input $x$와 가장 가까운 $K$개의 po..
study/IME654
2022. 7. 11. 11:28