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 |
- 백준
- 인공지능
- Cam
- Deep learning
- Unsupervised learning
- 딥러닝
- 설명가능한 인공지능
- 머신러닝
- AI
- SmoothGrad
- 코딩 테스트
- coding test
- 시계열 분석
- Interpretability
- grad-cam
- python
- 설명가능한
- keras
- meta-learning
- 기계학습
- Artificial Intelligence
- Class activation map
- GAN
- Explainable AI
- Score-CAM
- xai
- Machine Learning
- 메타러닝
- 코딩테스트
- cs231n
- Today
- Total
iMTE
이전 포스팅에서 정상성 시계열은 AR과 MA로 나타낼 수 있음을 설명하였다. 이번 포스팅에는 AR, MA 모형을 각각 설명하고 이 두개가 합친 ARMA 모형에 대해서 소개하고자한다. 먼저, AR의 기본적인 형태는 다음과 같다. $Z_t=\phi_1 Z_{t-1}+a_t$ 위 식에서 $\phi$는 정상성 조건을 만족시키며, $a_t$는 평균이 0, 분산이 $\sigma^2$을 갖는 white noise 이다. 위 식은 시차 1의 과거 값에 영향을 받음으로, AR(1) 모형이라고 부른다. 만약 시차 p인 AR(p) 모형은 다음과 같이 표현된다. $Z_{t}=\phi_1 Z_{t-1}+\phi_2 Z_{t-2}+...+\phi_p Z_{t-p}+a_t$ 위 식에서 $\phi$는 정상성 조건을 만족시킨다. 간단..
보통 시계열분석에서 stationary 시계열을 다루며, non-stationary 시계열은 여러 변환 (예, differentiation) 들을 통해서 정상적인 시계열으로 바꾼 이후에 분석을 한다. 1) 강정상성, strong stationary ($Z_1,...,Z_m$)과 ($Z_{1+k},...,Z_{m+k}$) ($k>1$)이 동일한 결합확률분포를 가질 때 strong stationry 시계열이라고 한다. 2) 위의 조건을 만족할 경우 다음 성질을 갖는다. (1) $E[Z_t]=\mu, t\geq 1$ (2) $Var[Z_t]=\sigma^2=\gamma (0), t\geq 1$ (3) $Cov[Z_t,Z_{t-k}]=Cov[Z_{t+k},Z_t]=\gamma (k), t\geq 1$ (자기공분..
모델을 만들어 낸 다음에, 모델의 예측 성능을 평가하기 위해서는 예측성능 척도가 필요하다. 먼저, error는 다음과 같이 정의가 된다. $$e_{t,1}=X_{t+1}-f_{t,1}$$ 위 식에서, $X_{t+1}$은 t+1 시점의 실제 값이고, $f_{t,1}$은 t 시점에서 모델에 의해 예측된 t+1 값을 의미한다. 만약 여러개의 관측치에 대한 예측 오차를 계산할 때, 4개의 예측 성능 척도가 사용된다. 1) Mean squared error (MSE) 2) Root mean squared error (RMSE) 3) Mean absolute deviation (MAD) 4) Mean absolute percent error (MAPE) 위 척도는 다음과 같은 수식으로 나타낼 수 있다. $MSE=\..
논문 제목 : Grad-CAM Guided Channel-spatial Attention Module for Fine-grained Visual Classification 논문 주소 : https://arxiv.org/abs/2101.09666 Grad-CAM guided channel-spatial attention module for fine-grained visual classification Fine-grained visual classification (FGVC) is becoming an important research field, due to its wide applications and the rapid development of computer vision technologies. Th..
논문 제목 : Ablation-CAM: Visual Explanations for Deep Convolutional Network Via Gradient-free Localization 논문 주소 : https://openaccess.thecvf.com/content_WACV_2020/html/Desai_Ablation-CAM_Visual_Explanations_for_Deep_Convolutional_Network_via_Gradient-free_Localization_WACV_2020_paper.html WACV 2020 Open Access Repository Ablation-CAM: Visual Explanations for Deep Convolutional Network via Gradient-..