Deep Learning Specialization 筆記
Introduction
Lecture 1 - Neural Networks and Deep Learning
Week1 Introduction to Deep Learning
Week2 LogisticRegression as a NeuralNetwork
Week2 Python and Vectorization
Week3 Shallow Neural Network
Week4 Deep Neural Network
Lecture 2 - Improving Deep Neural Networks
Week1 DataSets, Regularize, Optimization
Lecture 3 - Structuring Machine Learning Projects
Week1: ML Strategy(1)
Lecture 4 - Convolutional Neural Networks
W1: Foundations of CNN
W2: Case studies
W2: Practical advices for using Convnets
W3: Detection Algorithms
W4: Special Applications - Face recognition & Neural style transfer
Face Recognition
Neural Style Transfer
1D and 3D Generalizations
Lecture 5 - Sequence Models
Week1: Recurrent Neural Networks
Week2: Word Embedding
Week3: Seq2seq models
一般NN with sliding window 和 RNN 的區別在哪
Powered by
GitBook
一般NN with sliding window 和 RNN 的區別在哪
in my opinion
雖然一般NN 使用 sliding window,仍然可以學到前後關聯性,然而 RNN 可以選擇記住一些重要的 pattern,並且到了需要的時間點才使用這些 pattern,而這些時間點是不固定的,若使用一般NN,很可能學到固定的時間點。
相較於 NN,RNN 和 CNN 應該有點類似,有在不同 data 位置共享權重的特性,而這點也是 NN 辦不到的(嗎)。
results matching "
"
No results matching "
"