GithubHelp home page GithubHelp logo

ccs-lab / workshop_2018winter_deeplearning Goto Github PK

View Code? Open in Web Editor NEW
0.0 2.0 6.0 2.29 MB

Resources for the Deep Learning study (2018 Winter) with DeepLearningZeroToAll.

License: MIT License

Jupyter Notebook 27.81% HTML 72.19%
deeplearning-notebooks jupyter-notebooks

workshop_2018winter_deeplearning's Introduction

README

This repository is to study Deep Learning with people in CCS-Lab, in 2018 Winter (January - February 2018). All the materials in the repository can be used in addition to the Sung Kim's lectures and its codes.

Week 1 (2018-01-15)

Week 2 (2018-01-22)

  • Lectures
    • Logistic (Regression) Classification
      • Hypothesis 함수 소개
      • Cost 함수 소개
      • TensorFlow에서의 구현
    • Softmax Regression (Multinomial Logistic Regression)
      • Multinomial 개념 소개
      • Cost 함수 소개
      • Lab 1: TensorFlow에서의 구현
      • Lab 2: TensorFlow에서의 Fancy한 구현
  • Materials

Week 3 (2018-01-29)

  • Lectures
    • ML의 실용과 몇가지 팁
      • 학습 rate, Overfitting, 그리고 일반화 (Regularization)
      • Training/Testing 데이타 셋
      • Lab 1: TensorFlow에서의 구현 (학습 rate, training/test 셋으로 성능평가)
      • Lab 2: Meet MNIST dataset
    • 딥러닝의 기본 개념과, 문제, 그리고 해결
      • 딥러닝의 기본 개념: 시작과 XOR 문제
      • 딥러닝의 기본 개념2: Back-propagation과 2006/2007 '딥'의 출현
      • Lab: Tensor Manipulation
  • Materials

Week 4 (2018-02-13)

  • Lectures
    • Neural Network 1: XOR 문제와 학습방법, Backpropagation (1986 breakthrough)
      • XOR 문제 딥러닝으로 풀기
      • 특별편: 10분 안에 미분 정리하기
      • Deep Network 학습시키기 (backpropagation)
      • Lab 1: XOR을 위한 TensorFlow Deep Network
      • Lab 2: TensorBoard로 Deep Network 들여다보기
    • Neural Network 2: ReLU and 초기값 정하기 (2006/2007 breakthrough)
      • XSigmoid보다 ReLU가 더 좋아
      • Weight 초기화 잘해보자
      • Dropout과 앙상블
      • 레고처럼 넷트웍 모듈을 마음껏 쌓아 보자
      • Lab: 딥러닝으로 MNIST 98% 이상 해보기

Week 5 (2018-02-19)

  • Lectures
    • Convolutional Neural Networks
      • ConvNet의 Conv 레이어 만들기
      • ConvNet Max pooling 과 Full Network
      • ConvNet의 활용 예
      • Lab 1: TensorFlow CNN의 기본
      • Lab 2: TensorFlow로 구현하자 (MNIST 99%)
      • Lab 3: Class, tf.layers, Ensemble (MNIST 99.5%)

Week 6 (2018-02-26)

  • Lectures
    • Recurrent Neural Networks
      • NN의 꽃 RNN 이야기
      • Lab 1: RNN의 기본
      • Lab 2: Hi Hello RNN Traning
      • Lab 3: Long Sequence RNN
      • Lab 4: Stacked RNN + Softmax Layer
      • Lab 5: Dynamic RNN
      • Lab 6: Time Series RNN

Useful Resources

workshop_2018winter_deeplearning's People

Contributors

globalaisummit avatar harhimpark avatar hjingit avatar jaeyeongyang avatar jiminsun avatar youngahn avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.