The data is exactly the same as in the previous exercise. However, now you should learn a Bayesian classifier from it.
Transform the data into categorial form with suitable domain. Prefer very small domains, which contain only 2-4 attribute values. Often binary (0/1) data produces the most accurate model for small datasets. In any case, the final results FR should have only two values 0=failed and 1=passed.
a) Design a Bayesian network classifier for predicting course final results. Thus FR is the root node, which is updated by Bayes rule, given other attributes (excluding exam points). You can use either a Naive Bayes classifier or a more complex Bayesian network, but prefer simple structures.
b)Select randomly 1/5 of data as your test set and use the rest as a training set. Learn the model parameters from training data (you can use some tool, if you wish, or make a small program) and test it with the test set. Calculate the classification rates in the test set.