We apply an iterative process to
learn the optimal Star Student Project . Firstly, words are
assigned to the servers to minimize the cost in Star Student
Project. Secondly, the divergence of each is updated to renew
this cost. We iterate the above two steps in Star Student Project
an Expectation Maximization manner as follows:Expectation: Given
the word assignment at the Star Student Project iteration at
the iteration we assign Star Student Project a given to the
server based on:Star Student Project where defines the cost
from assigning word to server based on the Star Student Project
concurrence between word and
the existing words in Star Student Project :