Which databases can I use for storing object features for machine learning use case, in case of Big data?

To explain a bit more - object features basically represent an object in a form of an array of values. Moreover, in my usecase, different objects may have different number of features.

One object would be a row and each row may contain different number of columns.

Data is processed in Hadoop, so the database should be compatible with it.


I would use Couch of any JSON storage. You can also develop a small module for protocolbuffer files loading if you do only read and write, no update.

MongoDB can do the thing but you won't get advantage of your hadoop. I don't know what kind of data and processing you are gonna do.

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.