makeitrain

-- This is an application using machine learning to predict the future movement of the FTSE100 index. Each day, it will generate it's prediction for the FTSE100 for the next day and the accuracy of its prediction.

Front-page

How To Use

download miniconda
https://repo.continuum.io/miniconda/Miniconda3-latest-MacOSX-x86_64.sh

install miniconda
bash Miniconda3-latest-MacOSX-x86_64.sh

if you are using ZSH add miniconda path to ~/.zprofile
export PATH="$HOME/miniconda3/bin:$PATH"

conda env create -f makeitrain.yml
source activate makeitrain


to update the environment
conda env update -f makeitrain.yml

User Stories

EPIC

As a user
So I can make it rain
I want to predict the movement of the FTSE 100 Index

NOT SO EPIC

As a user
So I can know to buy or short
I want to receive an output of up or down

As a user
So I have confidence in the prediction
I want to see the percentage accuracy displayed

As a user
So I can minimize my investment risk
I want the prediction to have a minimum accuracy of 75%

Pending Improvements

Writing and implementing our own algorithm
Improve feature selection to increase our accuracy score
Process of getting daily data so result page is automatically updated not entirely functional (YET!)
Refactor party - namely for api.py
Reorganising repo and removing all unnecessary data files

Contributors

How to use

- Download miniconda
  https://repo.continuum.io/miniconda/Miniconda3-latest-MacOSX-x86_64.sh

- Install miniconda
  bash Miniconda3-latest-MacOSX-x86_64.sh

- If you are using ZSH add miniconda path to ~/.zprofile
  export PATH="$HOME/miniconda3/bin:$PATH"
- And then source the file the enable the changes
  source ~/.zprofile

- Create a new conda environment
  conda env create makeitrain

- Activate the newely created environment
  source activate makeitrain

- Install the needed dependecies
  pip install -r requirements.txt

Simo72 / makeitrain