As Deepti says, it is likely that your code does not generalize to the graders’ tests and therefore is not correct. Because the tests inside the running Lab are not the same as those of the grader.
I also noted that after this exercise, there is more cell that seems to be inaccurate between the text and the expected Cell results,
I mean in the non graded parts after exercise 4
That surely could be problem but what I am seeing a grader feedback not found error tells the autograder when wanted to grade the naive bayes, couldn’t find markers that detects grade function header or code markers ###START AND END CODE HERE### could be missing or deleted, added or edited anything outside the code markers.
no matter what your accurate output could also be reason.
it is probably time to check your codes, so click on my name and then message me screenshot of the grade function codes you are failing..
Make sure not to post any codes here as it is against code of conduct.
while computing p_spam for ham and spam, your codes are incorrect, for both the codes line you need to use class frequency with its respective class(i.e. ham or spam) over total emails.
so you using prob_email_given_spam or prob_email_given_ham to calculate p_spam and p_ham is incorrect.
Also remember in your prob_email_given_spam and ham codes you need to use parameter name=variable name, to avoid local/global variable recall conflict. Notice there is Parameters heading, which has the variable names, so for word_frequency=word_frequency need to followed, instead just mentioning the parameters name)
save emicon.
