Programming

Back to Homework page

References for programming homeworks:
- 深度學習快速入門—使用 TensorFlow，博碩，2017.
  - 原文書：G. Zaccone, Getting Started with TensorFlow, Packt, 2016. (PDF)
- Tensorflow tutorials 神经网络教学教程，周莫烦，44 部Youtube影片。網站說明文字。
Create "single" and "separate" report page for each programming assignment. Write the report page with
Submit the web address of your report page to Google Classroom.

P1: TensorFlow Basics

Goal: Understand what TensorFlow is, and install the TensorFlow 1.x.
Readings:
- "Chapter 1", Getting Started with TensorFlow.
- Tensorflow tutorials 神经网络教学教程，周莫烦，37 部Youtube影片之1-6部。
Install TensorFlow 1.x on your computer.
- You can install it on Linux, Mac OS, or Windows.
- You can choose to install CPU or GPU versions of TensorFlow.
There are two things you have to write in this report page
- Installation steps and practice codes (screen capture)
- Bug/difficulty you encounter, and the way you solve it.

P2: TensorFlow Basics II

Goal: Understand TensorFlow basic syntax, such as session, variable, placeholder.
Readings: Tensorflow tutorials 神经网络教学教程，周莫烦，Youtube影片之7-12部。
Report page: There are three things you have to write in this report page
- Execute all the sample codes in the 6 tutorial videos. Show and explain those sample codes.
- Bug/difficulty you encounter, and the way you solve it.
- Refer at least 5 web pages/books/materials for your learning of tensorflow in this homework. The book/youtube videos provided by the teacher can not be referred in your homework.

P3: Doing Math with TensorFlow

Goal: tensor data structure, fractals, gradients, and random number generation.
Readings
- "Chapter 2 Doing Math with TensorFlow", Getting Started with TensorFlow, G. Zaccone, Packt, 2016. 中文譯本：「第二章用TensorFlow求解數學問題」，深度學習快速入門─使用TensorFlow，G. Zaccone著/傅運文譯，博碩文化，2017。
Report page: There are three things you have to write in this report page

P4: Starting with Machine Learning

Goal: Implementation of linear regression, classifier(kNN for MNIST), and k-means clustering.
Readings
- "Chapter 3 Starting with Machine Learning", Getting Started with TensorFlow, G. Zaccone, Packt, 2016. 中文譯本：「第三章機器學習簡介與應用」，深度學習快速入門─使用TensorFlow，G. Zaccone著/傅運文譯，博碩文化，2017。
- TensorFlow Getting Started: Getting Started With TensorFlow, MNIST For ML Beginners.
Report page: There are three things you have to write in this report page
Option: TensorFlow SVM implementation for MNIST
- Apply both Linear SVM and Kernel SVM for MNIST.
- Please find SVM sample codes from: TensorFlow Course and Tutorial codes, GitHub. You can also find TensorFlow SVM code by yourself. However, you have to provide the references of your referred codes.

P5: Introducing Neural Networks

Goal: Implementation of perceptron, logistic regression, MLP, and function approximation.
Readings
- "Chapter 4 Introducing Neural Networks", Getting Started with TensorFlow, G. Zaccone, Packt, 2016. 中文譯本：「第四章類神經網路簡介」，深度學習快速入門─使用TensorFlow，G. Zaccone著/傅運文譯，博碩文化，2017。
- Tensorflow tutorials 神经网络教学教程，周莫烦，Youtube影片之13-22部。
- ConvnetJS: demo of toy 2d classification with 2-layer neural network
- TensorFlow Playground
Report page: There are three things you have to write in this report page
- Execute all the sample codes in the chapter. Show and explain those sample codes. Some extension and/or modification of the sample codes are welcome.
- Bug/difficulty you encounter, and the way you solve it.
- Refer at least 5 web pages/books/materials for your learning of tensorflow in this homework. The book/youtube videos provided by the teacher can not be referred in your homework.

P6: Deep Learning

Goal: Implementation of CNN and RNN.
Readings
- "Chapter 5 Deep Learning", Getting Started with TensorFlow, G. Zaccone, Packt, 2016. 中文譯本：「第五章深度學習」，深度學習快速入門─使用TensorFlow，G. Zaccone著/傅運文譯，博碩文化，2017。
- Quick complete Tensorflow tutorial to understand and run Alexnet, VGG, Inceptionv3, Resnet and squeezeNet networks, cv-tricks.com.
- Playground of CNN: ConvnetJS, MNIST demo
Report page: There are three things you have to write in this report page