Machine Learning: Here Comes Jack, the Reaper

Recently I have been moving a lot espeically in the state of mind and career choice. Later this July, I found myself in a small team which is extremely focusing to develop a machine learning engine for a game changing product. From that point , I started to know what are the entities currently running up in the machine learning fields. As I am having a bad Saturday, I would like to spend my time on something re-readable.

Here are the new things I have learned recently.


According to recent market validation, the Scala based Machine Learning server PreditionIO is getting a fair amount of attentions and community pick ups. It is based on popular technologies like Apache Spark, HBase and Spray.It is a complete solution for ML covering Algorithm Libraries like Mahout, Processing frameoworks like Spark and data store like HBase and Cassandra.

I have been using Hadoop for a binary ranking system which is implemented for my pet project Myanmar Archive. Computers can do everything we can do more repeatedly. We juse need to utilize this power and embrace the facts that we are a volatile species.

PreditionIO runs on a template based pattern for getting predictive entities. They have this place to share the templates and it is quiet colorful and exciting.


Neo4j is a pure graph database system and have a lot of use cases. This is my first experience with a graph database and it is hard to think differently from my old traditional SQL modeling.

My idea of modeling and relationships between entities are based on the foreign keys and table based thinking. I have a lot of fun digging this new thing and found it is quiet powerful for enterprise use cases like organisiational approval systems where you need a lot of entities to give decision on a specific request.

This is just a note of myself and sorry if this make your Saturday bad.

Happy Hacking

Copyright © 2016 - Tin Aung Linn. Powered by Octopress