This article is a collection of tutorials on implementing popular machine learning algorithms for TensorFlow. The goal is to make it easy for the reader to understand TensorFlow through clear and concise cases. These cases are suitable for beginners who want to implement some TensorFlow cases. This tutorial contains code that also contains notes and annotations.

Step 1: Tutorial guide for TF beginners

1: Tf beginners need to understand the preparation for entry

  • Introduction to Machine Learning notes:

aymericdamien/TensorFlow-Examples

  • MNIST Dataset Introduction Notes

aymericdamien/TensorFlow-Examples

2: Tf beginners need to know the basics of entry

  • Hello World

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

  • Basic operation

aymericdamien/TensorFlow-Examples
github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/1_Introduction/basic_operations.py

3: Basic models for TF beginners to master

  • Nearest neighbor:

aymericdamien/TensorFlow-Examples
github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/2_BasicModels/nearest_neighbor.py

  • Linear regression:

aymericdamien/TensorFlow-Examples
github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/2_BasicModels/linear_regression.py

  • Logistic regression:

aymericdamien/TensorFlow-Examples
github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/2_BasicModels/logistic_regression.py

4: NEURAL network that TF beginners need to try

  • Multilayer perceptron:

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

  • Convolutional neural network:

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

  • Cyclic neural Network (LSTM) :

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

  • Bidirectional cyclic neural Network (LSTM) :

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

  • Dynamic Recurrent Neural Network (LSTM)

Github.com/aymericdami…

  • Since the encoder

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

5: Tf beginners need to be proficient in practical technology

  • Save and restore the model

aymericdamien/TensorFlow-Examples
github.com/aymericdamien/TensorFlow-Examples/blob/master/examples/4_Utils/save_restore_model.py

  • Diagram and loss visualization

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

  • Tensorboard — Advanced visualization

Github.com/aymericdami…

6: TF beginners need to understand the basic operation of multi-GPU

  • Basic operations on multiple Gpus

aymericdamien/TensorFlow-Examples
Github.com/aymericdami…

7: Data set required by the case

There are cases where MNIST data sets are required for training and testing. When you run these cases, the data set is automatically downloaded (using input_data.py).

MNIST Dataset Notes: AymericDamien/Tensorflow-examples

Official website: MNIST Handwritten Digit Database, Yann LeCun, Corinna Cortes and Chris Burges

Step 2: Various types of cases, models, and data sets for TF novices

TFLearn TensorFlow

The following example comes from TFLearn, a library that provides a simplified interface to TensorFlow. There are lots of examples and pre-built operations and layers.

Use tutorial:

TFLearn quick start. Learn the basics of TFLearn through a specific machine learning task. Develop and train a deep neural network classifier.

TFLearn Address: TFLearn/TFLearn

Example: tflearn/tflearn

Pre-built operations and layers: index-TFLearn


Note: tflearn/tflearn

Basic models and data sets

  • Linear regression, using TFLearn to achieve linear regression

Github.com/tflearn/tfl…

  • Logical operators. Use TFLearn to implement logical operators

Github.com/tflearn/tfl…

  • The weight holds. Save and restore a model

Github.com/tflearn/tfl…

  • Fine tuning. Fine-tune a pre-trained model on a new task

Github.com/tflearn/tfl…

  • The use of HDF5. Use HDF5 to process large data sets

Github.com/tflearn/tfl…

  • Use the DASK. Use DASK to process large data sets

Github.com/tflearn/tfl…

Computer vision models and data sets

  • Multilayer perceptron. A multi-layer perception implementation for MNIST classification tasks

Github.com/tflearn/tfl…

  • Convolutional networks (MNIST). An implementation of convolutional neural network for classifying MNIST data sets

Github.com/tflearn/tfl…

  • Convolutional networks (CIFAR-10). An implementation of convolutional neural network for classifying CIFAR-10 data sets

Github.com/tflearn/tfl…

  • The network of networks. Network in Network implementation for classifying CIFAR-10 data sets

Github.com/tflearn/tfl…

  • Alexnet. Alexnet was applied to Oxford Flowers 17 classification task

Github.com/tflearn/tfl…

  • VGGNet. Apply VGGNet to Oxford Flowers 17 classification task

Github.com/tflearn/tfl…

  • VGGNet Finetuning (Fast Training). Use a pre-trained VGG network and constrain it to your own data for fast training

Github.com/tflearn/tfl…

  • RNN Pixels. Classify images using RNN (on a sequence of pixels)

Github.com/tflearn/tfl…

  • The Highway Network. Highway Network implementation for classifying MNIST data sets

Github.com/tflearn/tfl…

  • Taking Convolutional Network. Highway Convolutional Network implementation for classifying MNIST data sets

Github.com/tflearn/tfl…

  • Residual Network (MNIST). A Bottleneck Residual Network for MNIST Classification

Github.com/tflearn/tfl…

  • Residual Network (CIFAR – 10). A residual network for CIFAR-10 classification tasks

Github.com/tflearn/tfl…

  • Google Inception (v3). Google Inception V3 networking for Oxford Flowers 17 classification task

Github.com/tflearn/tfl…

  • Autoencoder. Autoencoder for MNIST handwritten digits

Github.com/tflearn/tfl…

Natural language processing models and data sets

  • Cyclic neural network (LSTM), using LSTM to IMDB emotion data set classification

Github.com/tflearn/tfl…

  • Bidirectional RNN (LSTM), applying a bidirectional LSTM to IMDB emotion data set classification task:

Github.com/tflearn/tfl…

  • Dynamic RNN (LSTM), using dynamic LSTM to classify variable-length text from IMDB data sets:

Github.com/tflearn/tfl…

  • City name generation, using LSTM network to generate a new US city name:

Github.com/tflearn/tfl…

  • Shakespeare manuscript generation, using the LSTM network to generate new Shakespeare manuscripts:

Github.com/tflearn/tfl…

  • Seq2seq, Seq2seq loop network teaching example:

Github.com/tflearn/tfl…

  • CNN Seq, a 1-D convolutional network was used to classify word sequences from IMDB sentiment datasets

Github.com/tflearn/tfl…

Reinforcement learning cases

  • Atari Pacman 1-step Q-learning, using 1-step Q-learning to teach a machine to play Atari games:

Github.com/tflearn/tfl…

Step 3: Other aspects for TF novices

  • Recommender-wide &Deep Network, teaching example of wide & deep Network in Recommender system:

Github.com/tflearn/tfl…

  • Spiral Classification Problem, TFLearn for Stanford CS231n Spiral Classification Problem:

tflearn/tflearn

  • Using the TFLearn layer with TensorFlow:

Github.com/tflearn/tfl…

  • The TFLearn trainer class is used to train any TensorFlow diagram:

Github.com/tflearn/tfl…

  • Bulit-in Ops, together with TensorFlow using TFLearn built-in operations:

Github.com/tflearn/tfl…

  • Fulllinks, along with TensorFlow, use TFLearn Summarizers:

Github.com/tflearn/tfl…

  • Variables, together with TensorFlow using TFLearn Variables:

Github.com/tflearn/tfl…

Reprinted from: dry | TensorFlow 55 classic case

The web links provided in this article are from the network, if you have any questions, please inform us.

Reprint please get the consent of the author BigQuant first!