Skip to content

Classification of time series data using neural nets

License

Notifications You must be signed in to change notification settings

joshemorris/timechange

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Timechange

Description

The purpose of the Timechange project is to design an open source software tool that simplifies as much as possible the task of sophisticated ML analysis of time series data. Moreover, Timechange will be unique in that it will serve as a one stop shop solution for data input, pre-processing, and transformation, as well as training and evaluation of state-of-the-art deep learning methods.

The focus with regards to the design of this system during the Spring 2017 Software Engineering course, will be on the implementation of features that will allow the system to generate images from transformed time series data, and then to use these images to train a deep learning model such as a convolutional neural network.

The focus with regards to the design of this system during the Fall 2017 Software Engineering course, will be:

  1. Converting Timechange into a web app through use of Flask
  2. Addition of the ability to train RNNs

Installation

  1. Clone this project using git

  2. install virtualenv

pip install virtualenv

Usage Instructions

  1. Create and run a virtual environment
#creates project
virtualenv my_project             

#activates projects
source my_project/bin/activate    
  1. Change to web directory

cd web

  1. Install requirements to run app

pip install -r requirements.txt

  1. start server

python manage.py runserver

  1. Enter the following address in your favorite browser

localhost:5000

Error: Too Many Open Files

You'll see this error if you try to upload more files than your system settings currently permit. This is not as big of an issue as it may seem at first.

Linux users should look here

Unix users should look here

About

Classification of time series data using neural nets

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 53.0%
  • HTML 30.7%
  • JavaScript 16.3%