Stock Clustering Part I

When come to building a well balanced stock portfolio, it is desired to have different mixture of stocks, potentially from different sectors to minimize the risk. Clustering is a great tool to discover these different types of clusters. You can use different features to categorize these stocks, eg. historical prices, returns, etc.

I know I could have used python or R to run the clustering. But as a big data explorer, I decided to use Mahout and Spark to do the job. I will post my findings and experiences of using these tools in the upcoming posts.

Here are the initial clusters found using the pre-clustering canopy algorithm in Mahout to determine the initial number of K clusters. I will perform the K-Means clustering algorithm next.

image(4)

image(3)

image(2)

image(1)

image

The tedious part of using Mahout clustering is you have to transform your data into sequence file format of mahout vector writables. I will share my codes soon.

Please check out my stock prediction model too 🙂 http://giantify.collective2.com/

To be continued…

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s