Fashion-MNIST Embedding Visualization

• Create date: 2018-11-19

Fashion-MNIST Embedding Visualization using TensorFlow Projector

Getting started

The CNN model

• The model summary

• The training loss and accuracy

Prerequisites

• numpy
• pickle
• scipy
• tensorflow==1.8
• keras
• sklearn
• Pillow
• matplotlib

Install the required packages by running the following command:

$ pip install -r requirements.txt

Usage

1. Clone this repository

$ git clone https://github.com/sthanhng/Fashion-MNIST-Embedding-Visualization

2. Run the following commands:

$ cd Fashion-MNIST-Embedding-Visualization
$ python -m http.server

3. Navigate to the localhost server http://127.0.0.1:8000/ to see the visualization.

Using t-SNE - [1792 iterations]

Reference

• TensorFlow Projector:

  • https://github.com/tensorflow/embedding-projector-standalone
  • https://projector.tensorflow.org/

• Fashion-MNIST with tf.Keras

  • https://medium.com/tensorflow/hello-deep-learning-fashion-mnist-with-keras-50fcff8cd74a