nBut recent instances
are more likely to reflect future
behavior
nA common technique
for that is to use exponential averaging
uS[n+1] = a T[n] + (1-a)
S[n] ; 0 < a < 1
umore weight is put on
recent instances whenever a >
1/n
nBy expanding this
eqn, we see that weights of past
instances are decreasing exponentially
uS[n+1] = aT[n] + (1-a)aT[n-1]
+ ... (1-a)^{i}aT[n-i] +
... + (1-a)^{n}S[1]
upredicted value of 1st instance S[1] is not calculated; usually set to 0 to give priority to to new processes
u