Consider a setting in which there are D diseases and a patient can either have or not have a particular disease d; = {0, 1}, for i = 1,...,D. Here di = 1 means that the patient has disease i'; d = 0 means the patient does not have disease 'i'. A patient may have more than one disease. There are a set of S symptoms the hospital can measure; if = Sj 1, for j = 1,..., then the patient has symptom j'; otherwise sj = 0 means that the patient does not have symptom j'. A simple disease-symptom network is given by S D p(1, ..., 8s, d,...,dp) = [[p(sj|d)][p(di) where d = (d,,dD) and j=1 i=1 p(s; = 1|d) = (w/d+bj) where (x) = 1/(1 + exp(x)). In the above wj is a vector of parameters relating symptom j to the diseases and bj is related to the prevalence of the symptom. The hospital provides the collection of parameters W and b, the prior disease probabilities p (with p(d; = 1) = p) and a vector s of symptoms for the patient, see SymptomDiseasePars.mat. Use Gibbs sampling (using a reasonable amount of burn-in and sub-sampling) to estimate the vector [p(d = 1|s),...,p(dd = 1|s)]