Notice
Recent Posts
Recent Comments
Link
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 |
Tags
- deep learning #machine learning #history
- recall
- FPS
- AI tech
- map
- 부트캠프
- Optimization
- objectdetection
- 네이버
- 일상
- 1x1 convolution
- 대학원생
- flops
- neural networks #linear #nonlinear #perceptrons
- evaluation
- generalization #overfitting # bootstrapping #bagging #boosting #stochastic gradient desent #mini-batch gradient descent #batch gradient descent
- confusion matrix
- precision
- momentum #adagrad #adam #early stopping #data augmentation #dropout #label smoothing #batch normalization
- IOU
- CNN #padding #stride #parameters
Archives
- Today
- Total
공부 일기장
Neural Networks & Multi-Layer Perceptron 본문
인간의 지능을 모방 but 똑같을 필요는 없다
과거의 새를 모방하고자 만든 비행기는 새와 똑같은가?
우리가 하늘을 날고 싶다고 해서 항상 새처럼 날 필요는 없다
neural networks가 시작은 인간의 뇌를 모방하고자 했지만 지금은 사실 많이 다르고 왜 잘됬는지는 수학적으로 분석해야 할 것이다.
Neural networks are fuction approximators that stack affine transformations followed by nonlinerar transformations.
- 비선형 연산이 반복적으로 일어나는 연산
Linear Neural Networks
NonLinear Neural Networks
multi-layer perceptrons
표현력을 극대화 하기 위해서는 nonlinear function이 필요
activation function
사실상 hidden layer가 하나만 있어도 충분한 표현력을 가짐
하지만 이는 단지 존재성만을 보이는 것이지 나의 모델이 그런 성질을 가진다는 보장은 하지 못함
'네이버 부스트 캠프 > DL basic' 카테고리의 다른 글
ILSVRC(modern models) (0) | 2023.03.06 |
---|---|
Convolutional Neural Networks (0) | 2023.02.27 |
Optimization (0) | 2023.02.27 |
Historical Review (0) | 2023.02.27 |
Comments