Convolutional Neural Network Walkthrough -Post 1 |Summer Research Summary

This post records the necessary links and excerpts to understand basic concepts regarding convolution neural network.

Preliminary:Building Platforms and Webtools

Building Platforms:

Linux System Useful commands.

Tensorflow
Open-source machine-learning platform. Dependent Language: Python [Free]
PyTorch
Succint and trouble-free building platform. Dependent Language: Python [Free]
C++
ML library written in C++. [Free]
Matlab / MatConNet
Using Matlab to build ML network. [May be paid]

Python-Tutorial
A comprehensive website for python beginners.

Useful webtools:

Github –Github-Installation / Github-Usage
Amazing collaborate platform, and good place to store your code.
Google VM – VM-Setup / GUI Support
Three hundred dollars worth of credits free to use. But GPU is not supported then.
Notepad++ –Notepad++ -Plugin
Free editing software with multi-functional hightlight.

Machine Learning

Machine Learning
Understand the basic concepts and category of machine learning.
Artificial Neural Network
Grasp some basic ideas about the composition of ANN.
Neural Network Playground
Have fun with the interactive design!
ANN
A more user-friendly introduction with multiple illustrations.

Visualizing Backpropagation
Understand backpropagation’s core issue: partial derivative.
Visualizing Backpropagation - Part Two
Step-by-step backpropagation with numbers.
Visualizing optimizers
You may refer to individual papers for detailed explanation.
MNIST using ANN
Auxiliary Notes. Be sure to try running real programs!
Tensorboard Readme
Visualizing and monitoring the training process.