In: Statistics and Probability
We can model the problem as Mrakov chain with states A, B, C and D.
The tratnsition probability from state A to state C or D is 1/2. The tratnsition probability from state A to state A or B is 0.
The transition probability from state B to state A is 4 times other transition probabilities (k) to C or D.
Thus 4k + k + k = 1
=> k = 1/6
The transition probability from state B to state A is 4/6 = 2/3 times and the transition probabilities to C or D is 1/6.
The tratnsition probability from state C to state D is 1. The tratnsition probability from state C to state A, B or C is 0.
The tratnsition probability from state D to state A, B, C or D is 1/4.
The transition probability matrix is,
The probability that mentioned dealer buys oil C when we know that oil B was purchased three weeks ago
= P(X3 = C | X0 = B)
= P(X0 = B, X1 = A, X2 = D, X3 = C) + P(X0 = B, X1 = C, X2 = D, X3 = C) + P(X0 = B, X1 = D, X2 = A, X3 = C) + P(X0 = B, X1 = D, X2 = B, X3 = C) + P(X0 = B, X1 = D, X2 = D, X3 = C)
= (2/3) * (1/2) * (1/4) + (1/6) * 1 * (1/4) + (1/6) * (1/4) * (1/2) + (1/6) * (1/4) * (1/6) + (1/6) * (1/4) * (1/4)
= 0.1631944