F21MP_2020 - Research-ML-BTC - H00359322

Introduction

As part of the course F21MP_2020-2021, Masters Project and Dissertation carried out at Heriot-Watt University for the MSc Data Science, the Research-ML-BTC project was carried out. This project is open-source and versioned on Git.

It is avaible on :

HWU Gitlab MACS (requires an HWU ID): Gitlab
The Github of the author Joseph Chartois: Github

Documentation

The documentation is generated with the Sphinx framework with the Read The Doc template. It is accessible via the file: /documentation/build/html/index.html

Pre-requisites (linux)

Git clones the repository:

git clone git@gitlab-student.macs.hw.ac.uk:jmmc2000/f21mp_2020-2021.git or
git clone git@github.com:JosephCHS/Research-ML-BTC.git

Enter the directory
Have Python 3 installed (default on recent OS like Ubuntu or Debian)
Install pip: sudo apt install -y python3-pip
Install virtual-env: python3 -m pip install --user virtualenv
Activate the virtual environment: source .venv/bin/activate
Install the prerequisites in the virtual environment: pip install -r source/requirements.txt

Usage

You can import modules from the project and use the classes and their methods. In addition, documentation is available: /documentation/build/html/index.html Examples of usage are written at the bottom of the files and commented to guide the user. These examples are also available bellow.

Fetch

dataset = Dataset()  
dataset.create_dataset()

Convert

convert = Arff()  
convert.generate_arff()  
convert.generate_arff_with_future()

Chart

Candlesticks

dataset = Dataset()  
dataframe_btc = dataset.get_btc_data()  
candlesticks = Candlesticks()  
candlesticks.display_candlesticks_chart(dataframe_btc)

Confusion_matrix

ConfusionMatrix(True)

Model

# Instantiate class  
machine_learning = MachineLearning()  
machine_learning.display_information()  
# Model sklearn  
models_sklearn = [  
    machine_learning.model_logistic_regression(),  
  machine_learning.model_svm(),  
]  
# Display models results sklearn  
for model in models_sklearn:  
    machine_learning.display_result(model)  
machine_learning.display_report_sklearn()  
# Model Keras  
machine_learning.model_lstm()  
machine_learning.model_cnn()  
# Model Pytorch  
machine_learning.model_bnn()

JosephCHS/Research-ML-BTC