Sequential drug decision problems in long-term medical conditions: a case Study of Primary Hypertension Eunju Kim ba, ma, msc

WHILE ~Finished; % Repeat until the following stopping rules are satisfied. IF itry >= MaxTries || Success >= MaxSuccess; IF T = MaxConsRej; Break; ELSE T = CoolSched (T); % Update T according to cooling schedule. END NewParam = Generator (Parent); % Generate a new solution. IF pertc == 30 % Every 30 repetitions, jump to another area. NewParam = NewParam + randi([-1000,1000]); END NewReward = EvModel(NewParam); % Evaluate the new solution. incNewReward = NewReward - OldReward; % If the new solution is better than the old solution, accept the new solution. IF (incNewReward > 1e-6) Parent = NewParam; OldReward = NewReward; % Otherwise, accept the new solution with a probability obtained from the Boltzman distribution. ELSE IF (rand > exp(-incNewReward/(k*T) )); Parent = NewParam; OldReward = NewReward; END END END % Decide the optimal solution and the reward where the optimal solution is used. OptSol = Parent; OptV = OldReward;

Figure ‎6.. Pseudo-code of the SA used for the hypertension SDDP model