Genuine Bayesians really consider restrictive probabilities as more essential than joint probabilities. It is anything but difficult to characterize P(A|B) without reference to the joint likelihood P(A,B). To see this note we can revamp the restrictive likelihood recipe to get: 

P(A|B) P(B) = P(A,B) 

be that as it may, by evenness we can likewise get: 

P(B|A) P(A) = P(A,B) 

It pursues that: 

image\BBNs0050.gif 

which is the purported Bayes Rule. 

It is entirely expected to consider Bayes rule as far as refreshing our conviction about a speculation An in the light of new proof B. In particular, our back conviction P(A|B) is determined by duplicating our earlier conviction P(A) by the probability P(B|A) that B will happen if An is valid. 

The intensity of Bayes’ standard is that much of the time where we need to figure P(A|B) things being what they are, it is hard to do so legitimately, yet we may have direct data about P(B|A). Bayes’ standard empowers us to register P(A|B) regarding P(B|A). 

For instance, assume that we are keen on diagnosing disease in patients who visit a chest center. 

Give An a chance to speak to the occasion “Individual has malignant growth” 

Give B a chance to speak to the occasion “Individual is a smoker” 

We know the likelihood of the earlier occasion P(A)=0.1 based on past information (10% of patients entering the center end up having malignant growth). We need to register the likelihood of the back occasion P(A|B). It is hard to locate this out legitimately. Be that as it may, we are probably going to know P(B) by considering the level of patients who smoke – assume P(B)=0.5. We are likewise liable to know P(B|A) by checking from our record the extent of smokers among those analyzed. Assume P(B|A)=0.8. 

We would now be able to utilize Bayes’ standard to register: 

P(A|B) = (0.8 ‘ 0.1)/0.5 = 0.16 

Consequently, in the light of proof that the individual is a smoker we change our earlier likelihood from 0.1 to a back likelihood of 0.16. This is an importance increment, yet it is still far-fetched that the individual has malignancy. 

The denominator P(B) in the condition is a normalizing consistent which can be registered, for instance, by minimization whereby 

image\BBNs0051_wmf.gif 

Henceforth we can state Bayes rule in another manner as: 

image\BBNs0052_wmf.gif