search

chengsoonong / digbeta

Active learning for Big Data
GNU General Public License v3.0
25 stars 14 forks source link

MEMM with k-th order Markov chain #98

Closed cdawei closed 7 years ago

cdawei commented 7 years ago

Train a maximum entropy Markov models with k-th order Markov chains, where k is the required number of POIs in query. This model treat trajectories conform to a specific query as samples from a (k-th order) Markov chain, and training is to estimate the parameters of Markov chains corresponding to all queries.

cdawei commented 7 years ago

It is not clear how to do the inference efficiently. Close the issue.