Okay, I see the data imbalance is a key factor here. I'm not sure if RMSE is the right choice, since that's more for regression tasks. I'm leaning towards F1 score or a weighted F-score to handle the imbalance.
I believe Key, Value, and TimeStamp are included because they are essential for identifying and storing data in HBase. Key Type is not necessary for the KeyValue format.
Tom
10 months agoKayleigh
10 months agoMammie
11 months agoElizabeth
11 months agoJeanice
11 months agoNoemi
11 months agoSherill
11 months agoOretha
12 months agoGlen
12 months agoJeanice
12 months agoDanica
12 months agoAlva
1 year agoShelba
1 year agoGertude
1 year agoJudy
1 year agoDeeanna
1 year agoPolly
1 year agoJames
1 year agoJanessa
1 year agoLashon
1 year agoAnnice
1 year agoShenika
1 year agoAretha
1 year agoMerissa
1 year agoJoye
1 year agoLynsey
1 year agoAmber
1 year agoFrank
1 year agoLaurel
1 year agoLuis
1 year agoHorace
1 year agoYun
1 year agoReita
1 year agoEvangelina
1 year ago