Analyzing Stock Correlations with Neo4j, PostgreSQL and Python Jupyter Notebook

In this video I will show how to visually analyze clusters within the stock market by correlating 3000 different symbols over the last 15 years, and showing the correlations using Neo4j Bloom. Firstly, I import the data into my PostgreSQL database, scaffold the symbols and dates, and impute missing values. I think upload the data to my local python Jupyter Notebook, and create a correlation matrix of all the symbols. Once I do that, I load the data into PostgreSQL where I create proper data referential integrity checks. I then upload the data into Neo4j, and after writing some basic Cypher queries, visualize it in Bloom application where I can play around with different clusters and symbols.

Leave a Comment