README.md

Tensorflow2로 배우는 머신러닝 [융합 신기술 교육 프로그램2(AA0020)]

이 git 저장소는 인하대학교 공학교육혁신센터 Tensorflow2로 배우는 머신러닝 [융합 신기술 교육 프로그램2(AA0020)] 강의 및 실습 내용을 요약정리한 내용을 담고 있습니다.

Index

1. Basic

   1. Python
   2. Numpy
   3. Pandas

2. KNN

3. Decision Tree

4. Regression Analysis

   1. Regression
      • Linear Regression
      • Least Square Method
      • Logistic Regression
      • Cost Function / Sigmoid Function
      • Gradient Descent Method
   2. Using Tensorflow
      • Linear Regression
      • Logistic Regression
      • MNIST Datasets
      • Categorical Encoding / Softmax Function
      • SGD / BGD
      • Epoch / Batch

5. Neural Network

   1. Neural Network
      • Neuron Cell / Perceptron
      • Neural Network
      • Hidden Layer
      • Backpropagation
   2. Using TensorFlow
      • Gradient Vanishing Problem
      • ReLU / LeakyReLU / ELU Function
      • Optimizers
        • Xavier / He Initialization
        • Dropout / Batch Normalization
        • Momentum / Nesterov Momentum
        • AdaGrad / RMSProp / Adam

6. Convolutional Neural Network

   1. CNN
      • Convolutional Layer
      • Pooling Layer
      • Toy Image
   2. MNIST CNN
   3. MNIST CNN DNN
   4. Example: Fashion MNIST
   5. Example: CIFAR10 Dataset
      1. CIFAR10 w/ Dropout (approx. 74%)
      2. CIFAR10 w/ Dropout, Batch Normalization (approx. 79%)
      3. CIFAR10 w/ Dropout, Batch Normalization, Kernel Regularizer (approx. 84%)
      4. CIFAR10 w/ Dropout, Batch Normalization, Kernel Regularizer, ImageDataGenerator (approx. 88%)
   6. Example: CatsVsDogs Dataset
      • README
      0. Split Data Set
      1. CNN_Train
      2. CNN_Test
      3. ImageGenerator_Train
      4. ImageGenerator_Test
      5. VGG16_Train
      6. VGG16_Test

7. Auto Encoder

8. Recurrent Neural Network

   1. RNN
      • Time Sequence Forecasting
      • RNN
        • one to one
        • one to many
        • many to one
        • many to many
        • Multiple Layer RNN
      • Example: Character RNN
      • TimeDistributed Layer
      • Embedding Layer
      • Projection Layer
   2. RNN
      • Gradient Vanishing in RNN
      • Long Short Term Memory
      • Stacked RNN
   3. Example: Stock Data
   4. Example: IMDB Dataset
   5. Word2Vec
   6. Seq2Seq
      • Attention

9. Appendix

   1. VAE
   2. Image / Video Preprocessing