14. Main Problem solving Areas - Classification
Algorithm Single machine MR Spark
Naive Bayes/
Complementary Naive
Bayes
Random Forest
Multilayer Perception
19. Goods and bads
● Several algorithms implementations ready to use
● Well documented java API
● More robust when compared to weeka
● Startup overhead when compared to Spark MLIB
● API target for programmers rather than data scientists
● Extensible API