This is a list of ongoing features/changes that I wish to add/make to the software:

• High priority:

  • Write documentation
    • Instructions for running on localhost: #30
    • Instructions for running on cluster: #31
    • Library documentation: #32
    • Spark integration examples: #28
  • Publish package to Sonatype: #33

• Medium priority:

  • Unify all parameters in the default configuration: #34
  • Automatic CI performance benchmark: #35

• Low priority (future work):

  • Implement sparse representations
  • Fault-tolerance through model replication
  • Instantaneous failover using DHT (similar to PS-lite)