https://github.com/marsbroshok/tensorflow-rnn-events-prediction

Tensorflow Recurrent Neural Network (RNN) model to analyse Time Series in GDELT News dataset to predict future events.
https://github.com/marsbroshok/tensorflow-rnn-events-prediction

Last synced: 3 months ago
JSON representation

Tensorflow Recurrent Neural Network (RNN) model to analyse Time Series in GDELT News dataset to predict future events.

• Host: GitHub
• URL: https://github.com/marsbroshok/tensorflow-rnn-events-prediction
• Owner: marsbroshok
• License: mit
• Created: 2016-06-28T13:53:35.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2017-07-19T09:09:46.000Z (over 7 years ago)
• Last Synced: 2024-07-19T22:48:00.297Z (4 months ago)
• Language: Jupyter Notebook
• Size: 516 KB
• Stars: 30
• Watchers: 3
• Forks: 13
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Tensorflow RNN to Events Prediction

**[NOTE]**: *This notebook was made with [Tensorflow v.0.8.0](https://github.com/tensorflow/tensorflow/releases/tag/v0.8.0) and code is not compatible with the newest release of Tensorflow. For the moment I don't have time to upgrade the code so you can use notebook more as an illustration of GDELT dataset and time series analysis.*

Tensorflow Recurrent Neural Network (RNN) model to analyse Time Series in GDELT News dataset to predict future events.

This Google Datalab notebook shows how to create a simple RNN model to perform analysis of GDELT news events dataset.

# Installation

To run this notebook, please follow these steps:

* Create Datalab project in Google Cloud Platform: https://cloud.google.com/datalab/docs/quickstart (do not worry about payment options, default quota is enought to run this demo notebook for free).

* Upload notebook [GDC Meetup Paris GDELT-skflow-RNN](https://github.com/marsbroshok/tensorflow-rnn-events-prediction/blob/master/GDC%20Meetup%20Paris%20GDELT-skflow-RNN.ipynb) to your Datalab envirement.

* Enjoy!

# Additional info

Please, check [presentation slides](https://docs.google.com/presentation/d/1EoUpLG11NDkKg00ay-YKo_RooQP9LY2RRt2RPFemdF8/edit?usp=sharing) based on this notebook.

_______________

Alexander Usoltsev, Cirruseo (Paris), 2016 (c)