[PYTHON] Summary of pages useful for studying the deep learning framework Chainer

Introduction

For those who want to use the deep learning framework Chainer for the first time and those who want to do image processing, language processing, etc. using Chainer, I have compiled various pages about Chainer. A brief summary of each page is also provided. I hope it will be useful for those who have too much information and do not know what to look for.

In the basic version, a page using the sample program that comes with Chainer, In the practical edition, we will mainly introduce pages that use MLP, RNN, and CNN for image classification, image recognition, and natural language generation.

For those who are new to neural networks and Chainer

Chainer installation method / environment settings

-Chainer Official Website -[Install CHAINER 1.5.1 on UBUNTU 14.04](https://daichiahl.wordpress.com/2016/01/15/chainer-1-5-1 installed on ubuntu14-04 /) -Neural network starting with Chainer

• Introduction to chainer -How to use Chainer and its application to natural language processing -Introduction of Machine Learning Library Chainer -[Introduction to Machine Learning] Until you run the sample code with chainer -Image classification using the Deep Learning framework Chainer Part 1

Reference book on deep learning

Deep Learning (Machine Learning Professional Series)

Deep Learning (Supervised by The Japanese Society for Artificial Intelligence)

Reference book on implementation using Python / Chainer

Deep Learning from scratch-the theory and implementation of deep learning learned in Python

Practical deep learning by Chainer

Basic edition

Handwriting recognition (mnist)

[Machine learning] I will explain while trying the deep learning framework Chainer.

How to get sample programs from the installation of Chainer, explanation of functions, etc. Quite detailed.

If you go back using chainer, it fits a little

Explanation of program changes and problems when using example / mnist to perform regression problems.

I had a regression problem with chainer

Program changes when doing regression problems using example / mnist.

Implementation of Convolutional Neural Network by Chainer

Implement mnist classification using Convolutional Neural Network.

Implementation of Multilayer Perceptron by Chainer

You can see the results when learning with some changes in example / mnist.

Image classification using the Deep Learning framework Chainer Part 2, [Part 3](http: // humansandcomputers.blogspot.jp/2015/08/deep-learning-chainer-3.html), [Part 4](http://humansandcomputers.blogspot.jp/2015/08/auto-image-classification-w-deep- learning-framework-chainer-4.html)

Detailed explanation of example / mnist. A description of the functions implemented in chainer.

Language model (ptb)

Read example \ ptb

The functions of example / ptb / train_ptb.py are explained in detail with comments.

Sentence generation using the model learned by ptb

Described the code to generate sentences using the model learned using example / ptb / train_ptb.py.

I tried to explain the sample code for creating a recurrent neural language model using Chainer

It explains the GPU driver settings and sample code from the installation of chainer. The version of chainer may be a little old.

Play with Chainer's ptb sample

Described the code to generate a sentence using the model learned using train_ptb.py.

Image recognition (imagenet)

Implementing Convolutional Neural Network

From the ease of ImageNet dataset to the explanation of Convolutional Neural Network.

Image classification with chainer, a deep learning framework from PFN (Neural network 1 with chainer)

Image classification using flickr style dataset.

Practical edition

Multilayer perceptron

Implementing a feedforward neural network with Chainer and classifying documents

Create one hot vector for each word to classify documents as positive or negative

I tried super-easy linear separation with Chainer

Determine if you are obese using your height (cm), weight (kg), and chest circumference (cm)

Let's learn neural network with chainer (Neural network with chainer 2)

Learning logical operators XOR and AND

Learning XOR

Learning the logical operator XOR

[Chainer] Learning XOR with Multilayer Perceptron

Learning the logical operator XOR

Recurrent neural network chainer-char-rnn Code that implements a character-level language model in Chainer

Learning with RNN to output literary text

Obtained Osamu Dazai's novel data from Aozora Bunko and learned the language model. Text generation is performed using the learned model.

Chainer, RNN and Machine Translation

Encoder-Implementation of translation model using Decoder.

[Evangelion] Try to automatically generate Asuka-like lines with Deep Learning

Learn RNNs using anime dialogue data.

Convolutional Neural Network

[Chainer] Document classification by convolutional neural network

Use the distributed representation vector of words learned in word2vec to classify documents as positive or negative.

Classifying anime faces by deep learning

Described the data preprocessing and model explanation from the method of acquiring anime face data.

Try to classify CIFAR-10 with Chainer

Classify images into 10 classes using the CIFAR-10 dataset.

Cainer's recognition of general objects in CIFAR-10 (1)

Classify images into 10 classes using the CIFAR-10 dataset. The explanation of the code is easy to understand.

in conclusion

If you have a recommended web page, please let me know.

Change log

-** Posted the first draft (2016/04/06) ** -** Added summary of some sites (2016/04/07) ** -** Added a summary of some sites (2016/04/10) ** -** Updated Implementation (Multilayer Perceptron) (2016/04/10) ** -** Updated language model (ptb) (2016/04/14) ** -** Updated Convolutional Neural Network (2016/07/03) ** -** Introduction updated (2016/12/10) ** -** Added summary of some sites (2016/12/10) ** --Updated from time to time ...