Missing feature values need to be addressed prior to the model development phase of the

CRISP$-$DM methodology to avoid training on incomplete data. This task assesses your ability to

navigate the complexities of constructing a data pipeline to transform partially erroneous raw

data to knowledge and evaluate the effectiveness of the proposed solution. The task will require

you to identify a suitable publicly available dataset$/$s$,$ for which there is previous research that

addresses the missing feature value problem. Propose an approach to use the Na$$ve Bayes

classifier to address the missing feature value problem in the context of categorical feature

values. Implement this approach on a publicly available dataset$/$s and report its performance in

the context of a classification problem. Compare the effectiveness of this approach against a

baseline imputation approach that uses the mode value, on two different machine learning

models. Justify and discuss your findings using appropriate metrics and relate your findings to

previous research on data imputation using the same dataset$/$s$.$ Present your findings in a

written report and a video presentation. State and motivate any assumptions or scope adopted

during the task