Machine Learning to Deep Learning_2 Tutorial code: https://github.com/leejaymin/TensorFlowLecture 2017. 02. 03 Jemin Lee (leejaymin@cnu.ac.kr) Hompage: https://leejaymin.github.io/index.html 역사 알고리즘 약간은 깊이 있는 이해 그것을 통한 실제 구현 과 응용 시스템 Area에서의 Deep-Learning

Table of Contents Fundamental Machine Learning (1일차) Linear Regression: Gradient Descent Algorithm (optimization) Logistic Regression (Single Neuron=Perceptron): Sigmoid (Logistic function), Convexity, Cross Entropy, Decision Boundary Multiple Perceptron (Hidden Layer): Backpropagation algorithm Deep Neural Network Breakthrough (2-3일차) Rebirth of Neural Network, renamed DNN TensorFlow Basic DNN, ReLU, Pre-training, Dropout Convolutional Neural Network (CNN) How to apply DNN into real world problem (4일차) Use-case: smarttention 2016

Deep Learning Framework

왜 TensorFlow 인가?

Large-Scale Deep Learning With TensorFlow, JeffDean 2016

Large-Scale Deep Learning With TensorFlow, JeffDean 2016

Large-Scale Deep Learning With TensorFlow, JeffDean 2016

Architecture Large-Scale Deep Learning With TensorFlow, JeffDean 2016

Portable & Scalable Large-Scale Deep Learning With TensorFlow, JeffDean 2016

How is TensorFlow used at Google? Recognizing Images with Inception Voice Recognition Smart reply in inbox by Gmail Large-Scale Deep Learning With TensorFlow, JeffDean 2016

실습: Basic TensorFlow 역사 알고리즘 약간은 깊이 있는 이해 그것을 통한 실제 구현 과 응용 시스템 Area에서의 Deep-Learning

Graph, Node, Edge 선언부와 실행부가 다름 코드 위치 “앞으로 어떻게 동작 할 것이다”라는 계획을 Graph로 표현한 것이 TensorFlow 이다. Operation: 동작을 정의 Node: Operation 정의 포함 Edge: node와 node를 연결함. 코드 위치 0.2.Basic/Basic Tutorial.ipynb

실습2: MNIST 역사 알고리즘 약간은 깊이 있는 이해 그것을 통한 실제 구현 과 응용 시스템 Area에서의 Deep-Learning

출처: 모두를 위한 머신러닝 시즌1, 김성훈 교수님

Coursera, Machine Learning, Andrew Ng

Diagnosing bias vs. variance High Bias problem == underfitting High variance problem == overfitting Coursera, Machine Learning, Andrew Ng

Coursera, Machine Learning, Andrew Ng

Learning curves[1/3] Coursera, Machine Learning, Andrew Ng

Learning curves[2/3] Coursera, Machine Learning, Andrew Ng

Learning curves[3/3] Coursera, Machine Learning, Andrew Ng

Summarized bias vs. variance Coursera, Machine Learning, Andrew Ng

Coursera, Machine Learning, Andrew Ng

코드 위치 https://github.com/leejaymin/TensorFlowLecture/blob/master/4.MNIST/MNIST_Tutori al_DNN.ipynb

구현 내용 기본적인 MLP 구현 11-layer MLP with ReLU Pre-training and Dropout One hidden layer Testing accuracy: ~92% 11-layer MLP with ReLU ReLU, Sigmoid, Softmax Testing accuracy: ~97% Pre-training and Dropout Xavier init. and new regularization Testing accuracy: ~99% Convolutional Neural Network Testing accuracy: ?%