In this project, we will explore K-Means clustering in unsupervised learning method that identifies separate price action clusters on candlestick data. In our example, we're applying K-Means Clustering on SPDR Gold Shares(gold-backed ETF) to visualise candlestick clusters chart. We will also create the cluster matrix that describes the probability of cluster i to cluster j, which is useful for forming predictive trading strategies based on today's cluster identification and prediction of subsequent cluster. The output of the candlestick clusters and cluster matrix are as shown as below:
- Open your terminal and clone the repository, then navigate to the the project folder.
git clone https://github.com/AndyTKH/K-Means-Clustering.git
cd K-Means-Clustering
- Open the notebook to view the project.
jupyter notebook k_means_gold.ipynb
- Simply close the terminal window to exit Jupyter Notebook.