michaelrdionne / module_10_challenge

Module_10_Challenge

Unsupervised Machine Learning: A novel approach to investment portfolios based on cryptocurrencies

---

Building a crypto-based financial prototype using:

• Jupyter Notebook

• Python

• Pandas

Evaluate crypto performance over different periods

Techniques used:

• Data Normalization with Standard Scaler

• Clustering with K-Means & PCA

• HV PLot for Comparative Visualizations

---

Technology Used

---

Installation Guide

1. The easiest way to install JupyterLab is to download the Anaconda application.

• To install Juypter Lab with pip:

$ pip install jupyterlab

• Then install Jupyter Notebook with pip:

$ pip install notebook

2. SciKit can be installed with pip:

$ pip install -U scikit-learn

• Or using the conda command

$ conda create -n sklearn-env -c conda-forge scikit-learn
$ conda activate sklearn-env

3. To install Pandas create a new (Python 3) notebook in Jupyter Notebook and then in the prompt enter:

$ import pandas

4. Recommended way to install hvPLot is using the conda command

conda install -c pyviz hvplot

• Or using PyPI

pip install hvplot

---

Pandas Usage

Pandas has many features, which was used in the code creation, and can be found at the link above.

hvPlot Usage

hvPlot has many features, which was used in the code creation, and can be found at the link above.

---

Contributors

---