Let, there will be N events say E1, E2, E3, … E(N)
And I have n number of sequences with a class(can be multiclass)/regression
LSTM or any NN cant be used as there are no huge sequences for each class… Also computation power has to be good like < 2 mins training time.
For multiclass classfication I tried using seqlearn python library(which uses hmm) and its working good. But for regression problem I dono how to proceed further.?
My data will look like:
X1[‘Event1’, ‘Event2’, ‘Event3’] —> Y[C1] / regression
X2[‘Event2’, ‘Event1’, ‘Event5’, ‘Event2’, ‘Event1’, ‘Event5’] —> Y[C2] / regression